Anatomy of Omicron neutralizing antibodies in COVID-19 mRNA vaccinees

SARS-CoV-2 vaccines, administered to billions of people worldwide, are mitigating the effects of the COVID-19 pandemic, however little is known about the molecular basis of antibody cross-protection to emerging variants, such as Omicron (B.1.1.529), and other coronaviruses. To answer this question, 276 neutralizing monoclonal antibodies (nAbs), previously isolated from seronegative and seropositive donors vaccinated with BNT162b2 mRNA vaccine1, were tested for neutralization against the Omicron variant and SARS-CoV-1 virus. Cross-neutralizing antibodies were isolated from 100% of seropositive and 20% of seronegative vaccinees. Only 14.2% and 4.0% of tested antibodies neutralized the Omicron variant and SARS-CoV-1 respectively. These nAbs recognized mainly the SARS-CoV-2 receptor binding domain (RBD) and targeted class 3 and class 4 epitope regions on the SARS-CoV-2 spike protein. Antibodies targeting class 1/2 epitope regions only rarely showed cross-neutralization activity. Cross-protective antibodies derived from a variety of germlines, the most frequents of which were the IGHV1-58;IGHJ3-1 and IGHV1-69;IGHV4-1. Only 15.6% and 7.8% of predominant gene-derived nAbs elicited against the original Wuhan virus cross-neutralized Omicron and SARS-CoV-1 respectively. Our data provide evidence of the presence of cross-neutralizing antibodies induced by vaccination and map conserved epitopes on the S protein that can inform vaccine design.


Introduction
Since its rst appearance in December 2019, more than 370 million cases of SARS-CoV-2 infections were reported worldwide, with over 5.6 million deaths.Effective vaccines against the virus that rst appeared in Wuhan, China, have been developed with unprecedented speed.However, their ability to contain the global pandemic has been compromised by the inability to timely deliver vaccines to low-income countries and by the appearance of several antigenic variants which escaped the natural and vaccineinduced immunity [2][3][4] .The main variants that emerged so far, and listed as variants of concern (VoCs), are named B.1.1.7 (Alpha), B.1.351(Beta), P.1 (Gamma), B.1.617.2 (Delta) and B.1.1.529(Omicron) 5,6 .The latter one showed to be the most e cient in spreading into partially immune populations and in few months from its appearance have conquered most regions of the world 7,8 .The Omicron variant contains 37 mutations in the spike (S) protein, with up to 15 mutations in the receptor binding domain (RBD), the primary target for neutralizing antibodies (nAbs).Several reports have shown that the unprecedented number of mutations carried on the Omicron S protein drastically reduce up to 40-fold the neutralizing e cacy of sera from infected and vaccinated people and that this VoC can escape more than 85% of nAbs described in literature, including several antibodies approved for clinical use by regulatory agencies [9][10][11][12][13][14][15][16][17] .While serum activity and neutralization e cacy of selected mAbs against Omicron have been extensively reported, the functional and genetic anatomy of nAbs elicited in naïve (seronegative) and convalescent (seropositive) people immunized with the BNT162b2 mRNA vaccine remains to be explored.Taking advantage of our previous work 1 , we tested 276 human monoclonal antibodies able to neutralize the original SARS-CoV-2 virus isolated in Wuhan, for their ability to neutralize the Omicron variant and the distantly related SARS-CoV-1 virus.Our work unravels the genetic signature of cross-protective antibodies and mapped conserved sites of pathogen vulnerability on the S protein that can be used to design the next generation of COVID-19 vaccines.

Omicron effects on vaccine-induced nAbs
To understand the impact of SARS-CoV-2 Omicron on the antibody response, we evaluated the neutralization activity of 276 nAbs previously isolated from seronegative (n=52) and seropositive (n=224) donors immunized with the BNT162b2 mRNA vaccine (Fig. 2) 1 .While Omicron cross-neutralizing nAbs were identi ed in all seropositives, these antibodies were found only in one out of ve seronegatives (20%) (Extended Data Table 1).Only 1 out of 52 nAbs from seronegatives (1.9%) neutralized Omicron with a medium-low neutralization potency (Fig. 2a and c).Since only 1 nAb was identi ed from this group no statistical difference in the geometric mean (GM) 100% inhibitory concentration (IC 100 ) between the Wuhan and Omicron viruses was found in this group.Conversely, 38 out 224 nAbs from seropositives (16.9%) were able to neutralize Omicron.These nAbs showed a 3.16-fold decreased neutralization potency compared to the Wuhan virus showing a GM-IC 100 of 719.8 ng/mL (Fig. 2b-c).Overall, 39 Omicron nAbs were identi ed of which 38 (97.4%) targeted the S protein RBD and 1 (1.6%) recognized the NTD (Extended Data Fig. 2b, left bar).None of the 39 identi ed nAbs showed an extreme neutralization potency (IC 100 below 10 ng/mL) against Omicron.To identify immunodominant sites on the Omicron S protein a ow cytometry-based competition assay was performed.In our previous study, we found that 215/276 (77.9%) nAbs bound to the S protein RBD and the majority of these antibodies were competing with J08, which epitope spans between Class 1 and Class 2 regions, and S309, which target the Class 3 region 1 .In this work all 215 RBD targeting-nAbs were additionally tested by competition assay against Class 4 targeting mAb CR3022.Class 1/2, Class 3 and Not-competing nAbs were found in both seronegatives and seropositives, while Class 4 competing nAbs were found exclusively in seropositives (Extended Data Fig. 2a; Extended Data Table 2).In both groups, Class1/2 competing nAbs were the most abundant constituting 70.3% (n=26) and 64.0%(m=114) of all antibodies isolated from seronegatives and seropositives respectively (Extended Data Fig. 2a; Extended Data Table 2).The second most abundant group of mAbs were Not-competing (n=4; 10.8%) and Class 3-competing nAbs (n=47; 27.0%) for seronegatives and seropositives respectively.Finally, only 7 (3.4%)Class 4 competing nAbs were found exclusively in seropositives (Extended Data Fig. 2a; Extended Data Table 2).From the 215 tested nAbs, we found that only 20 out of 140 (14%) Class 1/2 targeting nAbs were able to neutralize Omicron, while up to 29% percent (15/51) of Class 3 nAbs showed neutralization activity against this variant (Fig. 2d-e).Finally, 14% (1/7) Class 4-competing nAbs were able to neutralize the Omicron virus (Fig. 2f).Among the 39 RBD-targeting Omicron nAbs, the majority recognized the Class 1/2 region (n=20; 51.3%) followed by Class 3 (n=15; 38.5%), while nAbs targeting Class 4 region (n=2; 5.1%) and Not-competing (n=2; 5.1%) were the least represented (Extended Data Fig. 2C, left bar).

Discussion
In this work we deeply characterized an extensive panel of vaccine elicited-neutralizing human monoclonal antibodies against the heavily mutated Omicron variant and the distantly related SARS-CoV-1 virus to understand the degree of cross-protection and to map conserved regions on the S protein.We found that only 14.2% and 4.0% of our antibody panel was able to neutralize the Omicron VoC and the SARS-CoV-1 virus respectively.Remarkably, from the group of seronegative people vaccinated with two doses of the BNT162b2 mRNA vaccine, we isolated only 1 of the 39 nAbs against Omicron while none of their antibodies was able to neutralize SARS-CoV-1.Cross-neutralizing antibodies were isolated almost exclusively from people that were vaccinated after infection highlighting once again the broad crossprotection conferred by hybrid immunity 1,26 .Despite these results, we observed that antibody germlines mainly involved in Omicron and SARS-CoV-1 cross-protection (IGHV1-58;IGHJ3-1 and IGHV1-69;IGHJ4-1) were used in both seronegative and seropositive vaccinees suggesting that a third booster dose in naïve people could enhance germline-maturation and induce a more broad and persistent antibody response 27 .Furthermore, a third booster dose could drive a nity maturation of poorly cross-reactive but predominant RBD-targeting B cell germlines elicited following SARS-CoV-2 Wuhan infection or vaccination, which were shown to persist for up to 6 months in the draining lymph nodes of vaccinated individuals 28 .These data suggest that primary immunization with two doses of vaccine in naïve people is not su cient to elicit a meaningful proportion of cross-neutralizing antibodies, and that this requires a secondary immune response that can be obtained by vaccinating previously infected people or by providing a third booster dose 29 .Further studies will be necessary to understand whether a booster dose in naïve people can elicit a hybrid immunity-like antibody response and to de ne the molecular basis of cross-protection in this population.In addition to this observation, we herein de ned the epitope regions that mediate crossprotection with Omicron and SARS-CoV-1.Indeed, we observed that, while the majority of nAbs against SARS-CoV-2 Wuhan virus recognize the tip of the receptor binding domain, which comprise antigenic Classes 1 and 2, the larger proportion of cross-neutralizing antibodies map in the lower portion of the RBD which comprises exclusively antigenic Classes 3 and 4.This information can support the design of next-generation COVID-19 vaccines broadly protective against current and future variants.

Transcriptionally active PCR expression of neutralizing antibodies
The transcriptionally active PCR (TAP) expression of neutralizing antibodies (nAbs) was performed as previously described 1,20 .Antibodies heavy and light chain vectors were initially digested using restriction enzymes AgeI, SalI and Xho.PCR II products were ligated using the Gibson Assembly NEB into 25 ng of respective human Igγ1, Igκ and Igλ expression vectors 30,31 .TAP reaction was performed using 5 μl of Q5 polymerase (NEB), 5 μl of GC Enhancer (NEB), 5 μl of 5X buffer, 10 mM of dNTPs, 0.125 μl of forward/reverse primers and 3 μl of ligation product, using the following cycles: 98 °C for 2 min, 35 cycles 98 °C for 10 s, 61 °C for 20 s, 72 °C for 1 min and 72 °C for 5 min.TAP products were puri ed, quanti ed using the Qubit Fluorometric Quantitation assay (Invitrogen), and used for transient transfection in Expi293F cell line following manufacturer's instructions.

SARS-CoV-2 authentic viruses neutralization assay
All SARS-CoV-2 virus neutralization assays were performed in the biosafety level 3 (BSL3) laboratories at Toscana Life Sciences in Siena (Italy) and Vismederi Srl, Siena (Italy).BSL3 laboratories are approved by a Certi ed Biosafety Professional and are inspected every year by local authorities.To evaluate the neutralization activity of identi ed nAbs against SARS-CoV-2 and B.1.1.529(Omciron) VoC a cytopathic effect-based microneutralization assay (CPE-MN) was performed 1,20 .Brie y, nAbs were coincubated with a SARS-CoV-2 viral solution containing 100 median Tissue Culture Infectious Dose (100 TCID 50 ) of virus for 1 hour at 37°C, 5% CO 2 .The mixture was then added to the wells of a 96-well plate containing a sub-con uent Vero E6 cell monolayer.Plates were incubated for 3-4 days at 37°C in a humidi ed environment with 5% CO 2 , then examined for CPE by means of an inverted optical microscope by two independent operators.All nAbs were tested at a starting dilution of 1:8, diluted step 1:2, and the IC 100 evaluated based on their initial concentration.Technical duplicates for each experiment were performed.In each plate positive and negative control were used as previously described 1,20 .

SARS-CoV-2 S protein competition assay
Competitive Flow cytometry-based assay was performed to characterize nAbs binding pro les to SARS-CoV-2 S-protein as previously described 1 .Brie y, magnetic beads (Dynabeads His-Tag, Invitrogen) were covered with His-tagged S-proteins, following manufacturers' instructions.Then, 40 mg/mL of beadsbound-S-protein were incubated with unlabeled nAbs for 40 minutes at RT.Following incubation, samples were washed with PBS and incubated with uorescently labeled Class 1/2 (J08-A647), Class 3 (S309-A488) or Class 4 (CR3022-A647) S-protein nAbs binders.Antibodies labelling was performed using Alexa Fluor NHS Ester kit (Thermo Scienti c).Following 40 minutes of incubation at RT, beads-antibodies mix was washed with PBS, resuspended in 150 μL of PBS-BSA 1% and acquired using BD LSR II ow cytometer (Becton Dickinson).Results were analyzed using FlowJo™ Software (version 10).Beads with or without S-protein incubated with labeled antibodies were used as positive and negative controls respectively.

SARS-CoV-1 S protein binding assay
Expi293F cells (Thermo Fisher) were transiently transfected with SARS-CoV-1 S-protein expression vectors (pcDNA3.3_CoV1_D28)using Expifectamine Enhancer according to the manufacturer's protocol (Thermo Fisher).Two days later, to exclude dead cells from analysis, Expi293F were harvested, dispensed into a 96-well plate (3x10 5 cell/well), and stained for 30 minutes at room temperature (RT) with Live/Dead Fixable Aqua reagent (Invitrogen; Thermo Scienti c) diluted 1:500.Following Live/Dead staining, cells were washed with PBS and incubated with nAbs candidates for 40 minutes at RT. Next, to identify the SARS-CoV-1 S protein mAbs binders, cells were washed and stained with the Alexa Fluor 488labelled secondary antibody Goat anti-Human IgG (H+L) secondary antibody (Invitrogen) diluted 1:500.After 40 minutes of incubation, labeled cells were washed, resuspended in 150 μL of PBS and analyzed using the BD LSR II ow cytometer (Becton Dickinson).Cells incubated with the SARS-CoV-1 nAb binder (S309) or incubated only with the secondary antibody were used as positive and negative controls respectively.Data were analyzed with FlowJo™ Software (version 10).
HEK293TN-hACE2 cell line generation HEK293TN-hACE2 cell line was by lentiviral transduction of HEK293TN (System Bioscience) cells as described in Notarbartolo S. et al. 32 .Lentiviral vectors were produced following a standard procedure based on calcium phosphate co-transfection with 3 rd generation helper and transfer plasmids.The transfer vector pLENTI_hACE2_HygR was obtained by cloning of hACE2 from pcDNA3.

SARS-CoV-1 neutralization assay
For neutralization assay, HEK293TN-hACE2 cells were plated in white 96-well plates in complete DMEM medium.24h later, cells were infected with 0.1 MOI of SARS-CoV-1 pseudoparticles that were previously incubated with serial dilution of puri ed or not puri ed (cell supernatant) mAb .In particular, a 7-point dose-response curve (plus PBS as untreated control), was obtained by diluting mAb or supernatant respectively ve-fold and three-fold.Thereafter, nAbs of each dose-response curve point was added to the medium containing SARS-CoV-1 pseudoparticles adjusted to contain 0.1 MOI.After incubation for 1h at 37°C, 50 µl of mAb/SARS-CoV-1 pseudoparticles mixture was added to each well and plates were incubated for 24h at 37°C.Each point was assayed in technical triplicates.After 24h of incubation cell infection was measured by luciferase assay using Bright-Glo™ Luciferase System (Promega) and In nite F200 plate reader (Tecan) was used to read luminescence.Obtained relative light units (RLUs) were normalized to controls and dose response curve were generated by nonlinear regression curve tting with GraphPad Prism to calculate Neutralization Dose 50 (ND 50 ).
Functional repertoire analyses nAbs VH and VL reads were manually curated and retrieved using CLC sequence viewer (Qiagen).Aberrant sequences were removed from the data set.Analyzed reads were saved in FASTA format and the repertoire analyses was performed using Cloanalyst (http://www.bu.edu/computationalimmunology/research/software/) 34,35 .The graph shows the neutralization potency (ND 50 ) of predominant gene derived nAbs against SARS1 (i).

Statistical analysis
Numbers and percentages of nAbs tested for each germline and neutralization ranges (black dotted lines) are denoted on each graph.j-k, violin plots show differences in the aminoacidic (aa) CDRH3 length and percentage of V gene somatic mutations for all IGHV1-58;IGHJ3-1 (j) and IGHV1-69;IGHJ4-1 (k) compared to Omicron and SARS1 nAbs respectively.
Supplementary Files

Figure 1 Distribution
Figure 1

Figure 2 Functional
Figure 2

Figure 3 Functional
Figure 3