Antibodies capable of binding to the spike protein of the coronavirus sars-cov-2

Abstract

The present disclosure relates to antibodies useful for the prevention, treatment and / or diagnosis of coronavirus infection and diseases and / or complications associated with coronavirus infection, including COVID-19. In particular, the present disclosure relates to antibodies capable of binding to the spike protein of the coronavirus SARS-CoV-2 and uses thereof.

Description

TECHNICAL FIELD

[0001] The present invention relates to antibodies useful for the prevention, treatment and / or diagnosis of coronavirus infections and diseases and / or complications associated with

[0002] coronavirus infections, including COVID-19. BACKGROUND

[0003] In December 2019, the first case of a severe viral acute respiratory syndrome named COVID-19 was reported. The virus spread rapidly worldwide, leading to a pandemic with more than 200 million confirmed infections and more than 4.4 million deaths in 12 months. The causative agent, SARS-CoV-2, is a beta coronavirus related to SARS-CoV-1 and MERS coronavirus, which also cause severe respiratory syndromes.

[0004] In the months since SARS-CoV-2 was identified as the causative agent of COVID-19, great progress has been made in our understanding of the disease and the virus. There are now a number of validated treatments, including dexamethasone and tocilizumab (Tocilizumab) and monoclonal antibodies (mAbs), which have been shown to be effective when used in prophylactic and therapeutic settings (Baum et al., 2020, Science 369, 1014-1018). Despite these advances, the pandemic is far from being under control, leading to successive waves of infections.

[0005] Coronaviruses have four structural proteins: the nucleocapsid, envelope, membrane, and spike (S) protein. The spike protein is the most important surface protein. It has an elongated, trimeric structure and is responsible for engaging the target cell and triggering fusion of the virus with the host membrane. The spike protein from both SARS-CoV-2 and SARS-CoV-1 uses angiotensin-converting enzyme 2 (ACE2) as a cell surface receptor. ACE2 is expressed in many tissues, including epithelial cells of the upper and lower respiratory tract.

[0006] The S protein consists of two subunits, S1 which mediates receptor binding and S2 which is responsible for virus and host cell membrane fusion. It is a dynamic structure that can transition to a post-fusion state through cleavage between S1 and S2 upon receptor binding or trypsin treatment. In some SARS-CoV-2 sequences, a furin cleavage site is inserted between the S1 and S2 subunits and mutations in the cleavage site attenuate disease in animal models. The S1 fragment occupies the membrane distal tip of S and can be subdivided into the N-terminal domain (NTD) and the receptor binding domain (RBD). While both regions are immunogenic, the RBD contains the interaction surface for ACE2 binding. Although the RBD is normally squished against the top of S2 down, it can swing up to engage ACE2. Monoclonal antibodies (mAbs) recognize one or both of the “up” and “down” conformations.

[0007] The S protein is relatively conserved between SARS-CoV-2 and SARS-CoV-1 (76%), but the degree of conservation is lower for the RBD and NTD (74% and 50%, respectively) than for the S2 domain (90%). The degree of conservation is much lower for MERS-CoV and seasonal human coronaviruses (19-21%). Overall, SARS-CoV-2 antibodies show limited cross-reactivity, even with SARS-CoV-1.

[0008] S is involved in viral attachment to target cells via its interaction with the S receptor binding motif (also known as the ACE-2 footprint), a 25 amino acid patch at the apex of the receptor binding domain (RBD) in the S1 fragment of the spike protein, to the cell surface expressed ACE2. Upon attachment, cleavage of S releases S1, allowing a major conformational change in S2 to expose the hydrophobic fusion loop to perform fusion of the viral and host cell membranes, releasing the viral genome into the host cell cytoplasm to initiate viral replication. Analysis of a large number of mAbs generated from SARS-CoV-2 infected individuals showed that mAbs bind to multiple epitopes on S1 and S2. Most mAbs generated against the original strain of SARS-CoV-2, while able to bind S with high affinity, showed little or no neutralization activity. Genomic surveillance of SARS-CoV-2 has identified thousands of mutations in the structural and non-structural proteins. However, at the end of 2020, viral variants were described that rapidly became the dominant strain locally and led to global spread and were designated as Variants of Concern (VoCs).

[0009] Alpha (B.1.1.7) was first identified in the UK and increased transmissibility. B.1.1.7 has 9 amino acid changes in the spike, including N501Y in the ACE2 interaction surface. Beta (501Y.V2, also known as B.1.351) was first reported in South Africa. Gamma (P.1, 501Y.V2) were first reported in Brazil and have 10 and 12 amino acid changes in the spike protein, respectively. Delta was first reported in India and has now spread globally, causing outbreaks in multiple countries. Omicron BA.1 was first reported in southern Africa in late November 2021 and spread to all parts of the world, becoming the dominant variant in many countries and almost completely replacing Delta.

[0010] A series of sublineages of Omicron have emerged, including BA.1.1, BA.2, BA.2.12.1, BA.2.75, and BA.4 / 5, which outcompete previous strains and become regional or global dominants. Over 30 mutations are found in the Omicron S protein, including 15 substitutions in the RBD, leading to increased transmissibility (Suzuki et al., 2022 “Attenuated fusogenicity and pathogenicity of SARS-CoV-2 Omicron variant.” Nature 603, 700-705) and widespread and substantial reduction in neutralizing antibody titers (Dejnirattisai et al., 2022 “SARS-CoV-2 Omicron-B.1.1.529 leads to widespread escape from neutralizing antibody responses.” Cell 185, 467-484 e415).

[0011] Omicron BA.2 was reported almost simultaneously with BA.1. The proportion of Omicron infections caused by BA.2 has been increasing in multiple countries and became the dominant sublineage in Denmark and India.

[0012] BA.1.1 (containing an additional R346K mutation in the RBD) once accounted for approximately 40% of global Omicron sequences, and approximately 35-60% in the UK and US (Iketani et al., 2022 “Antibody evasion properties of SARS-CoV-2 Omicron sublineages.” Nature 604, 553-556), but was soon overtaken by BA.2. As of August 2022, BA.2 (containing 8 unique substitutions in S, including 6 within the RBD, and lacking 13 mutations found in BA.1 (Nutalai et al., 2022)) has become the globally dominant strain. Recently, BA.2.12.1 has been identified in multiple countries and caused widespread regional outbreaks in North America (accounting for 58% of sequences as of May 25, 2022) (Del Rio and Malani, 2022, “COVID-19 in 2022 - The Beginning of the End or the End of the Beginning?” JAMA 327, 2389-2390).

[0013] It is now increasingly clear that BA.2 has a small transmission advantage over BA.1, although there is no evidence of increased disease severity. In early April 2022, two new Omicron lineages were reported in the Gauteng province of South Africa, named BA.4 and BA.5. BA.4 and BA.5 (with identical S sequences) became the dominant Omicron strains in Gauteng and triggered a new wave of infections in South Africa.

[0014] Since June 2022, BA.4 / 5 (with higher receptor binding affinity and significantly enhanced escape from antibody responses (Tuekprakhon et al., 2022 “Antibody escape of SARS-CoV-2 Omicron BA.4 and BA.5 from vaccine and BA.1 serum.” Cell 185, 2422-2433 e2413)) rapidly spread from South Africa to the globe and have now become the new globally dominant strains, with BA.5 in the ascendancy in many regions. These variants, particularly BA.5, now account for the majority of sequenced cases in many countries.

[0015] In early May 2022, a new Omicron sublineage emerged in India, designated BA.2.75. This strain later spread to many countries, including the United Kingdom, the United States, Australia, Germany, and Canada. However, the true prevalence of BA.2.75 is difficult to determine, as sequencing in many countries is sporadic and has been greatly reduced.

[0016] All these variants contain multiple mutations in S and include changes in the RBD, NTD, and in some cases the furin cleavage site between S1 and S2. The RBD mutations found in Alpha (N501Y), Beta (K417N, E484K, N501Y), Gamma (K417T, E484K, N501Y), and Delta (L452R, T478K) are located in or immediately adjacent to the ACE2 interaction surface, where they have the potential to modulate ACE2 interaction and disrupt neutralizing antibody binding. The increase in affinity for ACE2 interaction was mainly for Alpha, Beta, Gamma, and Delta (7, 19, 19, 2-fold, respectively) and can play a role in increasing viral transmissibility. Omicron contains an unprecedented number of mutations, concentrated in the spike (S) gene, which carries 30 substitutions, plus a 6-residue deletion and 3-residue insertion. Omicron BA.1 (RBD mutations G339D, S371L, S373P, S375F, K417N, N440K, G446S, S477N, T478K, E484A, Q493R, G496S, Q498R, N501Y, Y505H) contains the unique mutations S371L, G446S, and G496S, and in some isolates R346K (BA.1.1), while BA.2 carries S371F, T376A, D405N, and R408S. BA.3 contains no unique mutations relative to BA.1 and BA.2 and appears to be a fusion of the two, BA.1-like at the N-terminus and BA.2-like at the C-terminus by the mutation G496S.

[0017] Compared to BA.2, BA.2.75 contains multiple mutations in the S protein, including four substitutions in the NTD (W152R, F157L, I210V, and G257S) and four substitutions in the RBD: D339H, G446S, N460K, and R493Q.

[0018] Three new variants related to BA.2, BA.2.11, BA.2.12.1, and BA.2.13, have also been detected in multiple countries. These new variants contain single mutations L452R, L452Q, and L452M, respectively, compared to the BA.2 spike protein receptor binding domain (RBD) Figure 29). Of these, BA.2.12.1, first identified in New York, dominates in the US, accounting for about 58% of SARS-CoV-2 isolates as of May 25, 2022. While L452R was found in Delta and Kappa, L452Q was found in Lambda, L452M is new.

[0019] Given the physicochemical properties of the side chain of residue 452, BA.2.13 is expected to be a relatively modest change; the change of L to M will increase the size of the side chain, but it will still be hydrophobic. The change of L to Q in BA.2.12.1 will introduce some polar character, while BA.2.11 is the most radical, where the change of L to R will introduce a large basic amino acid.

[0020] Additional variants, BA.4 and BA.5, which have the same S sequence, appear to have evolved from BA.2. The sequences of BA.4 and BA.5 are highly related to that of BA.2, but contain additional mutations. Specifically, residues 69 and 70 of the NTD have been deleted (also found in Alpha, BA.1, and BA.3) and they contain two additional substitutions in the RBD: L452R (also found in Delta) and F486V. Finally, BA.4 and BA.5 lack the change of Q493R seen in BA.1 and BA.2, reverting to Q493 as in the Victoria / Wuhan strain. When looking at the RBD, BA.4 and BA.5 have the assembled mutations at all of the previously described positions in the VoCs Alpha (N501Y), Beta (K417N, E484K, N501Y), Gamma (K417T, E484K, N501Y), Delta (L452, T478K), with the only difference being E484A in BA.4 and BA.5 instead of E484K Beta and Gamma.

[0021] As of September 2022, a new variant related to BA.4 / 5 (designated BA.4.6) has emerged and expanded in the United States where BA.5 was dominant (as of September 10, 2022, at 87.5% prevalence, increasing from less than 2% of sequences in early July 2022 to more than 6% in mid-August 2022). Compared to BA.4 / 5, BA.4.6 contains two additional mutations in the spike protein (S), R346T in the RBD and N658S in the C-terminal domain. The R346T mutation has raised concerns about enhanced BA.4 / 5 antibody evasion, as the R346K mutation in BA.1.1 reduced serum neutralization and impaired the activity of many monoclonal antibodies (mAbs) compared to BA.1 (Nutalai et al., 2022). SARS-CoV-2 test kits using monoclonal antibodies have also been developed. Examples include lateral flow tests by, for example, Innova (SARS-CoV-2 Antigen Rapid Qualitative Test) and Quidel (Sofia 2 SARS Antigen FIA). However, these tests have been reported to be very inaccurate.

[0022] As of January 2023, more variants have emerged, such as BQ.1 and XBB, which carry up to 8 additional RBD amino acid substitutions compared to BA.2.

[0023] The structural-functional map of the panel of monoclonal antibodies (mAbs) isolated from infected cases has provided a fair understanding of the antigenicity of S and the mechanisms of neutralization. Most potent neutralizing antibodies bind at or very close to the footprint of ACE2 and act by blocking the ACE2 interaction, thus preventing cell attachment and infection. A second interaction site of potent mAbs is close to the N-linked glycan at position N343, exemplified by S309, which do not block ACE2 interaction but can act to destabilize the S trimer. A third group of potent mAbs binds to the N-terminal domain in S1, and its mode of action is currently unknown. Another RBD epitope of interest is located outside the ACE2 footprint, although the mAbs that bind here are not potent neutralizers, they can be effective in protecting in vivo (Huo et al., 2020; Sun et al., 2021; Yuan et al., 2020; Zhou et al., 2020).

[0024] Several new trends were observed during the evolution of Omicron after BA.5: i) the emergence of “second-generation” BA.2 variants (including derivatives of BA.5) - variants with long phylogenetic branch lengths, multiple antigenic mutations, and a lack of genetic intermediaries, such as BA.2.75, BJ.1, BS.1, BA.2.10.4, and BA.2.3.20 (van der Straten et al., 2022. Immunity 55, 1725-1731); and ii) accelerated antigenic drift, seen in both BA.5 (Tuekprakhon et al., 2022) and these second-generation BA.2 lineages, particularly in BQ.1 and BA.2.75 (https: / / nextstrain.org / nextclade / sars-cov-2 / 21L). Finally, recombination between two of these second-generation variants (BJ.1 and BM.1.1.1) resulted in XBB. Many of these variants show a high degree of convergent evolution in known antigenic RBD residues, and mutations are located in regions that can threaten neutralizing antibody binding, leading to further escape from protection against infection provided by vaccination or prior SARS-CoV-2 infection, including prior Omicron infection.

[0025] Currently, many lineages within the BA.2 and BA.5 clades are rapidly growing. Most notable is the high degree of convergent evolution, particularly at antigen RBD positions such as 346, 444, 446, 452, 460, 486, 490, and 494. These lineages include examples from the BA.4 / 5 clade (naturally containing L452R, F486V, and the back mutation R493Q), such as BA.4.6 and BF.7 (R346T), BA.4.7 (R346S), BQ.1 (K444T, N460K), and BQ.1.1 (R346T, K444T, N460K); from the BA.2.75 clade (naturally containing G446S, N460K, and the back mutation R493Q), BA.2.75.2 (R346T and F486V), BN.1 (R346T, K356T, F490S). There are also several other examples of second-generation BA.2 variants, such as BJ.1 (also known as BA.2.10.1.1; R346T, L368I, V445P, G446S, V483A, and F490V), BA.2.10.4 (G446S, F486P, S494P, and the R493Q back mutation), BS.1 (BA.2.3.2.1; R346T, L452R, N460K, G476S), BA.2.3.20 (K444R, N450D, L452M, N460K, E484R, and Q493R back mutations), and finally the BA.2.75 x BJ.1 recombinant XBB (which contains R346T, L368I, V445P, G446S, N460K, F486S, F490S relative to BA.2).

[0026] These second-generation BA.2 variants have become globally dominant, with BQ.1 alone accounting for 50% of infections as of December 27, 2022 (Past6M). https: / / cov-spectrum.org / explore / World / AllSamples / XBB.1.5 (XBB.1 + F486P) is rapidly expanding in North America.

[0027] Outside of RBD, the degree of convergent evolution is less, but still present. Many second-generation BA.2 variant lineages contain deletions or mutations in the NTD, often similar to those seen in VoCs, such as Δ~144 in BJ.1 and BA.2.10.4 (previously seen in Alpha and BA.1) and NSP12 G671S in BJ.1, BA.2.75, and BA.2.10.4 (previously seen in Delta).

[0028] All currently approved SARS-CoV-2 vaccines were designed to induce antibody (and T cell) responses against S and contain the S sequence found in the original ancestral strain. Therefore, there is particular concern whether S mutations in VoCs can cause immune escape, leading to vaccine failure or easy re-infection of previously infected individuals.

[0029] The broad mutational burden in Omicron S disrupts the activity of most mAbs to bind to the above three potent antibody binding sites (ACE-2 footprint, N343 glycan surrounding, and NTD). This can lead to a severe reduction or complete loss of serum neutralization capacity due to natural infection or vaccination, resulting in enhanced transmissibility and explosive spread of Omicron.

[0030] It is the object of the present invention to identify further improved antibodies useful for the prevention, treatment and / or diagnosis of coronavirus infections and diseases and / or complications associated with coronavirus infections, including COVID-19, especially the Omicron variant of concern (VoC) and yet to be identified variants with additional mutations in the ACE-2 footprint, RBD and / or NTD in the spike protein of SARS-CoV-2. SUMMARY

[0031] The inventors identified 28 human monoclonal antibodies (mAbs) recognizing the spike protein of SARS-CoV-2 (see Table 3). These antibodies show potent neutralizing activity against SARS-CoV-2. Some of the antibodies in Table 3 show potent neutralization against the hCoV-19 / Wuhan / WIV04 / 2019 strain and against SARS-CoV-2 strains from different lineages such as Victoria (Wuhan + S247R), Alpha, Beta, Gamma, Delta, Omicron (including Omicron BA.2.11, Omicron BA.2.12.1, Omicron BA.2.13, Omicron Omicron BA.2.3.20, Omicron BA.2.10.4, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, Omicron BA.2.75, BA.2.75.2, Omicron BA.3, Omicron BA.4.6, Omicron BA.4 / 5, Omicron BJ.1, Omicron BS.1, Omicron BN.1, Omicron XBB, and / or Omicron XBB.1 strains).

[0032] Many of the mAbs in Table 3 use a public V gene (a V gene that is common to most people). The inventors have previously shown that additional antibodies can be generated by swapping the light and heavy chains of antibodies in Tables 1, 2 and 3 that are derived from the same public V gene. Antibodies derived from the same public V gene provide particularly useful mixed chain antibodies.

[0033] In particular, the inventors have found that antibodies Omi02, Omi03, Omi12, Omi18, Omi28, Omi39 and Omi42 are particularly effective at cross-neutralising SARS-CoV-2 strains Victoria, Alpha, Beta, Gamma, Delta and Omicron.

[0034] Accordingly, the present application provides an antibody capable of binding to the spike protein of the coronavirus SARS-CoV-2, wherein the antibody comprises at least three CDRs of any one of the 28 antibodies in Table 3.

[0035] The present application provides an antibody capable of binding to the spike protein of the coronavirus SARS-CoV-2, wherein the antibody comprises at least three CDRs of antibody Omi12 or any one of the 27 antibodies in Table 3.

[0036] The present application also provides an antibody combination comprising two or more antibodies according to the present application.

[0037] The present application also provides an antibody combination comprising (a) an antibody of the present application; and (b) an antibody comprising at least three CDRs of an antibody in Table 1 or Table 2. For example, the antibody can comprise (i) at least four, five or all six CDRs of an antibody in Table 1 or Table 2; (ii) a heavy chain variable domain comprising or consisting of an amino acid sequence having at least 80% sequence identity to the heavy chain variable domain of an antibody in Table 1 or Table 2; (iii) a light chain variable domain comprising or consisting of an amino acid sequence having at least 80% sequence identity to the light chain variable domain of an antibody in Table 1 or Table 2; and / or (iv) a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having at least 80% identity to the heavy chain variable domain and light chain variable domain, respectively, of an antibody in Table 1 or Table 2.

[0038] The present application also provides one or more polynucleotides encoding an antibody of the present application, one or more vectors comprising said polynucleotide or a host cell comprising said vector.

[0039] The present application also provides a method for producing an antibody capable of binding to the spike protein of the coronavirus SARS-CoV-2, the method comprising culturing the host cell of the present application and isolating the antibody from the culture.

[0040] The present application also provides a pharmaceutical composition comprising: (a) an antibody or antibody combination of the present application, and (b) at least one pharmaceutically acceptable diluent or carrier.

[0041] The present application also provides an antibody, antibody combination or pharmaceutical composition of the present application for use in a method of treatment of the human or animal body by therapy.

[0042] The present application also provides an antibody, antibody combination or pharmaceutical composition of the present application for use in a method of treating or preventing a coronavirus infection or a disease or complication associated with a coronavirus infection.

[0043] The present application also provides a method of treating or preventing a coronavirus infection or a disease or complication associated with a coronavirus infection, the method comprising administering to the subject a therapeutically effective amount of an antibody, antibody combination or pharmaceutical composition of the present application.

[0044] The present application also provides a method of identifying the presence of a coronavirus or a protein fragment thereof in a sample, the method comprising (i) contacting the sample with an antibody or antibody combination of the present application, and (ii) detecting the presence or absence of an antibody-antigen complex, wherein the presence of an antibody-antigen complex is indicative of the presence of a coronavirus or a fragment thereof in the sample.

[0045] The present application also provides a method of treating or preventing a coronavirus infection or a disease or complication associated therewith in a subject, the method comprising identifying the presence of a coronavirus according to the method of the present application, and treating the subject with an antibody or combination according to the present application, an antiviral or anti-inflammatory agent.

[0046] The present application also provides the use of an antibody, antibody combination or pharmaceutical composition of the present application for the prophylaxis, treatment and / or diagnosis of a coronavirus infection or a disease or complication associated therewith.

[0047] The present application also provides the use of an antibody, antibody combination or pharmaceutical composition of the present application for the manufacture of a medicament for the treatment or prevention of a coronavirus infection or a disease or complication associated therewith. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 Generation of BA.2 sublineage of Omicron and a panel of Omicron mAbs. Figure 1Only the first 22 Omicron antibodies disclosed in Tables 13 and 14 (i.e. Omi02 to Omi35) were involved. (A) FRNT50 titres against Victoria and Omicron BA.1 from the donors used to generate the Omicron mAbs are shown. (B) FACS plots showing the sorting of B cells using full-length Omicron S. (C) The proportion of RBD and NTD binding antibodies found in the Omicron mAbs compared to the earlier pandemic mAbs. (D) Heavy and light chain variable genes used. (E) Somatic mutations found in the potent Omicron mAbs (FRNT50 < 100 ng / ml) compared to the earlier pandemic group.

[0049] Figure 2. Neutralisation curves using the Omicron mAbs. (A) Victoria, Alpha, Beta, Gamma, Delta and Omicron BA.1 viruses. (B) Neutralisation of Victoria, BA.1, BA.1.1, BA.2 and BA.3 pseudoviruses by the Omicron mAbs. (C) Neutralisation of Victoria, BA.1, BA.1.1, BA.2 and BA.3 pseudoviruses by the antibodies being developed for commercial use.

[0050] Figure 3 . Neutralisation of Victoria, BA.1, BA.1.1, BA.2 and BA.3 pseudoviruses. Neutralisation of Victoria, BA.1, BA.1.1, BA.2 and BA.3 pseudoviruses 28 days after the second and third doses of (A) AZD1222 (n=41), (B) BNT162b2 (n=20). (C) Live virus neutralisation assays on Victoria, Alpha, Beta, Gamma, Delta and Omicron viruses using sera obtained <14 days and >21 days after symptom onset. (D) Neutralisation of Victoria, BA.1, BA.1.1, BA.2 and BA.3 pseudoviruses by early and late sera. Geometric mean titres are shown above each column. Analysis was performed using Wilcoxon paired signed-rank test (A and B) and Mann-Whitney test (C and D) and two-tailed P values were calculated.

[0051] Figure 4. Pseudovirus neutralisation curves. Pseudovirus neutralisation curves of BA.1, BA.1.1, BA.2 and BA.3 against the earlier pandemic mAbs (B) Beta mAbs.

[0052] Figure 5 . With Figure 3Related neutralization titers of the indicated viruses. (A) Live viruses (B) Pseudoviruses. Geometric mean titers are shown above each column. Analysis was performed using Wilcoxon paired signed-rank test and two-tailed P values were calculated. (C) Pseudovirus neutralization curves of selected VH1-58 mAbs against Victoria and Iota (S477N) and control VH3-53 mAb 222.

[0053] Figure 6 Structure of BA.1 RBD with Omi-12 Fab. (A) Comparison of the Omi-12 and Beta-54 Fabs with BA.1 in two ternary complexes in the crystal asymmetric unit by superimposing the RBD (generated by fitting the high-resolution structures of BA.1 RBD, Omi-12, and Beta-54 to the lower-resolution ternary complex densities). The Fabs in one complex are in bright colors (cartoon depicting HC in red, LC in blue), while the other is in light colors. (B) Binding mode of Omi-12. (C) Close-up of the differences in binding of Omi-12 with Fab 253 to the early pandemic RBD (light blue) and Beta-47 with Beta RBD (light cyan). (D) The somatic mutation V53P helps to refold the H3 loop so that Q493R can be accommodated in Omi-12.

[0054] Figure 7 Pseudovirus neutralization assay of vaccine and BA.1 immune sera against BA.4 / 5. IC50 values of the indicated viruses using sera obtained from vaccinees at (A) 28 days after the third dose of AstraZeneca (AZD1222) (n=41), (B) 4 weeks after the third dose of Pfizer (BNT162b2) (n=20). Sera from volunteers who experienced breakthrough BA.1 infection were collected at (C) early after symptom onset, < 114 days (median 13 days) (n=12), (D) late after symptom onset, > 21 days (median 38 days) (n=16). Neutralization titers are compared to Victoria (early pandemic strain), BA.1, BA.1.1, BA.2, and BA.3. Geometric mean titers are shown above each column. Analysis was performed using Wilcoxon paired signed-rank test and two-tailed P values were calculated.

[0055] Figure 8 Pseudovirus neutralization assay against Omicron and commercial monoclonal antibodies. Neutralization curves of a panel of 28 monoclonal antibodies made from samples collected from vaccinees infected with BA.1. Titration curves of BA.1 are compared to BA.1, BA.1.1, BA.2, and BA.3. mAbs affected by L452R and F486L are indicated.

[0056] Figure 9 Comparison of S protein mutations of Omicron sublineages compared to BA.4 / 5. (A) Comparison of S protein mutations of Omicron BA.1, BA.1.1, BA.2, BA.3, and BA.4 / 5, with the NTD and RBD boundaries indicated. (B) Location of RBD mutations (gray surface, ACE2 footprint in dark green). Mutations common to all Omicron lineages are shown in white (Q493R, which is a back mutation in BA.4 / 5, is shown with a cross), mutations common to BA.1 and BA.1.1 are shown in cyan, mutations specific to BA.1.1 are shown in blue, and mutations specific to BA.2 are shown in magenta. Residue 371 (yellow) is mutated in all Omicron viruses, but differs between BA.1 and BA.2. The N343 glycan is shown as a rod with a transparent surface.

[0057] Figure 10 Surface plasmon resonance (SPR) analysis of the interaction between BA.2 or BA.4 / 5 RBD and selected mAbs. (A) Binding of BA.4 / 5 RBD was severely reduced compared to binding to BA.2, thus no accurate determination of binding could be made, as shown by a single injection of 200 nM RBD over a sample flow cell containing IgG Omi-31. (B-C; E-I) Sensorgrams (red: raw binding curve; black: fitted curve) showing the interaction between BA.2 or BA.4 / 5 RBD and selected mAbs, with kinetic data shown. (D) Affinity of BA.4 / 5 RBD for Omi-12 was determined using a 1:1 binding equilibrium analysis.

[0058] Figure 11 Interaction between mAbs and BA.4 / 5 mutation sites. (A) BA.1-RBD / Omi-31 (PDB 7ZFB), (B) BA.1-RBD / Omi-32 (PDB 7ZFE), (C) BA.1-RBD / Omi-25 (PDB 7ZFD), (D) BA.1-RBD / Omi-42 (PDB 7ZR7), (E) Wuhan-RBD / AZD8895 (PDB 7L7D), and (F) BA.1-RBD / Omi-3 (PDB 7ZF3) complexes overall structure (left panel) and interaction with BA.4 / 5 mutation sites (right panel). Residues are colored according to the following scheme: BA.4 / 5 mutation sites in BA.1-RBD are shown in magenta, BA.1-RBD residues in BA.1-RBD / Omi-3 complex are shown in cyan, and BA.1-RBD residues in BA.1-RBD / Omi-25 complex are shown in yellow. Omi-3 and Omi-25 are shown as stick models in white and light blue, respectively. The N343 glycan is shown as a rod with a transparent surface. (Right panel). In the left panel, the RBD is shown as a surface representation with the BA.4 / 5 mutation sites highlighted in magenta and the additional two mutation sites in BA.4 / 5 at 452 and 486 shown in cyan, the Fab LC is shown in blue and the HC in red strips. In the right panel, the side chains of the RBD, Fab HC and LC are drawn as grey, red and blue sticks, respectively. In (B), L452R is modelled to show the salt bridge (yellow broken stick) that can form to D99 of CDR-H3. (D) The Beta-RBD / Omi-42 complex shows that the Fab does not contact either of the two BA.4 / 5 mutation sites.

[0059] Figure 12 ACE2 RBD affinity. (A)-(D) SPR sensorgrams showing the binding of ACE2 to BA.4 / 5 RBD (A) compared to binding to ancestral (B), BA.1 (C) and BA.2 RBD (D). Data for ancestral, BA.1 and BA.2 have been previously reported (Nutalai et al., 2022). (E)-(G) Electrostatic surfaces: (E) From left to right are early pandemic, Delta and BA.1 RBD, (F) the open view of BA.2 RBD and ACE2 of the BA.2 RBD / ACE2 complex (PDB 7ZF7), and (G) BA.4 / 5 RBD (modelled based on the structure of BA.2 RBD). Diamonds on ACE2 and RBD show the areas of interaction.

[0060] Figure 13 Antigenic landscape. (A) Neutralisation data and model used to calculate the antigenic landscape in (B) (log titre values). Rows indicate sera collected from vaccinated volunteers or infected patients. Columns are challenge strains: in order are Victoria, Alpha, Delta, Beta, Gamma, BA.1, BA1.1, BA.2, BA.3 and BA.4 / 5. Values are coloured according to deviation from a reference value; reference values were calculated based on the mean of neutralisation titres of rows giving the highest values for each serum type. (B) Orthographic view of the antigenic landscape showing BA.4 / 5 in the context of previous VoCs and positions of BA.1, BA.1.1, BA.1 and BA.2 (calculated from pseudovirus neutralisation data). The distance between two positions is proportional to the reduction in neutralisation titre when one of the corresponding strains is infected by the serum challenged by the other. Figure 6 ACE2 / RBD affinity and antigenic landscape

[0061] Figure 14Neutralisation curves for VH1-58 mAb. Early pandemic mAb 253 (Dejnirattisai et al., 2021a) and Beta-47 (Liu et al., 2021b) neutralisation curves for pseudoviruses of Victoria and the panel of Omicron lineage constructs.

[0062] Figure 15 Surface plasmon resonance (SPR) analysis of the interaction between BA.2 or BA.4 / 5 RBD and selected mAbs. (A-F) Sensorgrams (red: raw binding curves; black: fitted curves) showing the interaction between BA.2 or BA.4 / 5 RBD and selected mAbs, with kinetic data shown. (G-K) Binding of BA.4 / 5 RBD was severely reduced compared to BA.2, so accurate determination of binding could not be made, as shown by a single injection of 200 nM RBD over a sample flow cell containing the indicated mAb.

[0063] Figure 16 Sequence changes in BA.2.75 compared to other Omicron sublineages. (A) Sequence alignment of BA.2.75 with Omicron sublineages Omicron BA.1, BA.1.1, BA.2, BA.3 and BA.4 / 5. Boundaries of NTD and RBD are labelled. (B) Surface representation of mutated residues in BA.2.75 RBD compared to BA.2 RBD. Positions mutated in BA.2 RBD are shown (grey surface with dark green ACE2 footprint) and residues mutated in BA.2.75 are shown in orange and labelled.

[0064] Figure 17 Pseudovirus neutralisation assay of BA.2.75 by vaccine and BA.1 and BA.2 immune sera. IC50 values for the indicated viruses using sera obtained from vaccinees 28 days after the third dose of vaccine. (A) Pfizer BNT162b2 (n=22). (B) AstraZeneca AZD AZD1222 (n=41). (C, D) Sera from volunteers who experienced vaccine breakthrough infection with BA.1 (n=16) or BA.2 (n=23). (EC) IC50 values for single RBD point mutations in an insert BA.2 pseudovirus using Pfizer BNT162b2 sera (n=22). Geometric mean titres are shown above each column. Analysis was performed using Wilcoxon paired signed-rank test and two-tailed P values were calculated.

[0065] Figure 18ACE2 / RBD affinity. SPR sensorgrams showing the binding of ACE2 to BA.2.75 RBD using either ACE2-Fc (A) or biotinylated ACE2 as ligand (B) compared to binding to RBD of BA.2 (C), BA.4 / 5 (D), Alpha (E) and BA.2+R493Q (F). Data for BA.2, BA.4 / 5 and Alpha have been previously reported in Nutalai et al., 2022, Tuekprakhon et al., 2022 and Dejnirattisai et al., 2022 respectively.

[0066] Figure 19 Pseudovirus neutralization assay against monoclonal antibodies. (A) Neutralization curves for a panel of 28 mAbs made from samples taken from vaccinees infected with BA.1. Titration curves for BA.2.75 are compared to Victoria, BA.1, BA.1.1, BA.2 and BA.4 / 5. IC50 titres are shown in Table 22. (B) Pseudovirus neutralization assay developed for mAbs for human use. IC50 titres are shown in Table 23. Data for Victoria, BA.1, BA.1.1 and BA.2 and BA.4 / 5 are for comparison and taken from Tuekprakhon et al., 2022

[0067] Figure 20 Structure of BA.2.75 RBD / ACE2 complex. (A) Overall structure of BA.2.75 RBD / ACE2 complex. ACE2 is shown as a green ribbon and RBD as a surface with mutations common to BA.2 highlighted in magenta and different mutations highlighted in orange. (B) Comparison of the interface of BA.2.75 RBD (grey) and ACE2 (green) to that of BA.2 and ACE2 (both light orange). Close-ups show the interactions of Q496R and Q493 (R493 in BA.2) with ACE2.

[0068] Figure 21Interactions between mAbs and BA.75 mutation sites. (A) Omi-3 (PDB, 7ZF3) and Omi-18 (PDB, 7ZFC) binding modes in front and back views of the superimposed RBD complexed with omicron BA.1 RBD. The RBD is shown as a gray surface with the mutations shared between BA.2 and BA.2.75 colored in magenta and the four mutations different between the two colored in cyan. The VH and VL are shown as ribbons and colored in red and blue for Omi-3 and light blue and light orange for Omi-18, respectively. (B) Interactions between N460 of the RBD and CDR-H2 of the Fab. (C) Contacts between R493 of the RBD and CDR-H3 of the Fab. In (B) and (C), the RBD related to Omi-3 is in gray, the RBD related to Omi-18 is in cyan, and the colors of the Fab are the same as in (A). (D) AZD1061 binding to ancestral SARS-CoV-2 RBD (PDB, 7L7E) and (E) contacts between G446 of the RBD and CDR-L2 of the Fab. (E) AZD8895 binding to ancestral SARS-CoV-2 spike protein RBD (PDB, 7L7E) and (F) contacts between Q493 of the RBD and CDR-H2 of the Fab. In (D)-(F), the rendering and coloring of the RBD are the same as in (A), HC is red, and LC is blue.

[0069] Figure 22 . Antigenic landscape. (A) Orthographic view of the antigenic landscape showing BA.2.75 in the context of the locations of previous VoCs and BA.1, BA.1.1, BA.1, and BA.2 (calculated from pseudovirus neutralization data). The distance between two locations is proportional to the reduction in neutralization titer when one of the corresponding strains is infected by the other upon serum challenge. No scale is provided because the figure is a projection of a three-dimensional distribution, but the changes can be calculated by making the following comparisons: (i) BA.1 vs. BA.2, a 2.93-fold reduction, and (ii) BA.2 vs. BA.4 / 5, a 3.03-fold reduction. (B) Same as (A) but including only the Omicron sublineage and early pandemic viruses to allow a more accurate projection of this subset into three-dimensional space. Note that the responses of these viruses to all sera are included in the calculation.

[0070] Figure 23 . Pseudovirus neutralization assays against monoclonal antibodies. (A) Neutralization curves for a panel of 28 monoclonal antibodies made from samples taken from BA.1 infected vaccinees. The titration curve for the single mutation of BA.2.75 in the BA.2 backbone is compared to BA.2 and BA.2.75. IC50 titers are shown in Table 24.

[0071] Figure 24 Surface plasmon resonance (SPR) analysis of the interaction between BA.2 or BA.2.75 RBD and selected mAbs. (A) Binding of Omi-29 (IGHV3-53) to BA.2.75 RBD is severely reduced compared to binding to BA.2, as shown by a single injection of 1 mM Omi-29 Fab over sample flow cells containing biotinylated BA.2 or BA.2.75 RBD. (B) Binding of Omi-36 (IGHV3-66) to BA.2.75 RBD is severely reduced compared to binding to BA.2, as shown by a single injection of 0.2 mM BA.2 or BA.2.75 RBD over sample flow cells containing Omi-36 in IgG format. (C-H) Sensorgrams (red: raw binding curves; black: fitted curves) showing the interaction between BA.2 or BA.4 / 5 RBD and selected mAbs, with kinetic data shown.

[0072] Figure 25 Neutralization of BA.2.75 by multiple panels of convalescent sera collected from infections with historical variants. Neutralization titers of the indicated sera against BA.2.75 and the indicated pseudoviruses. Data from Tuekprakhon et al., 2022, except for BA.2.75.

[0073] Figure 26 Primers used for site-directed PCR mutagenesis of BA.2.75 RBD. Site-directed PCR mutagenesis was performed using the BA.2 spike construct as a template. The indicated primers were used to introduce the D339H, G446S, N460K, and R493Q mutations.

[0074] Figure 27BA.2.11, BA.2.12.1 and BA.2.13 were characterized by pseudovirus neutralization assays, surface plasmon resonance and structural analysis. (a), (b) IC50 values for the indicated viruses obtained using sera obtained 4 weeks after the third dose of (a) AstraZeneca AZD1222 (n=41), (b) Pfizer BNT162b2 (n=18). (c) Neutralization titers taken from sera from vaccinated volunteers who suffered a breakthrough BA.1 infection. Comparison with previously reported neutralization titers for Victoria, BA.1, BA.1.1, BA.2 and BA.4 / 5 in Tuekprakhon et al. (2022). Geometric mean titers are shown above each column. Analysis was performed using Wilcoxon paired signed-rank test and two-tailed p values were calculated. (d-g) SPR sensorgrams (red: experimental binding curves; black: fitted curves) showing the binding of ACE2 to the RBD of BA.2.11 (e), BA.2.12.1 (f), BA.2.13 (g) compared to the binding to BA.2 RBD (h), where the kinetic data are shown. Data for BA.2 RBD were reported in Nutalai et al. (2022). (h-m) Crystal structure of the BA.2.12.1 RBD / Beta-27 / NbCl complex. (h) The overall structure is shown as C a traces with the RBD (grey), the Beta-27 HC (red) and the LC (blue) and NbCl (yellow). The C a of residues L452Q, F486 and Q493R (L, F and R in BA.2, R, V and Q in BA.4 / 5) are shown as spheres. (i) Comparison of the Beta-27 binding mode in the BA.2.12.1 RBD / Beta-27 / NbCl (RBD is shown as surface representation, HC is red and LC is blue), BA.4 / 5 RBD / Beta-27 / NbCl (cyan, PDB 7ZXU) and Beta RBD / Beta-27 (green, PDB 7PS1) complexes by superimposing the RBDs. Apart from the flexible N- and C-terminal regions of the RBD, the N-terminus of the Fab HC and CDR-H1, the a2 helix of the RBD, the 371-375 loop and the G446 loop appear significantly different. CDR-L3 has a dual conformation in the BA.4 / 5 RBD complex, while it has a single conformation in the other two complexes (i). The HC N-terminus and CDR-H1 contacting residue 486 of the RBD are different in the Beta and BA.4 / 5 RBD complexes, the latter containing the F486V mutation. These differences can be caused by the contacts with the symmetry-related C1 nanobody, which is shown as grey bonds in (j). (k) Structural differences at the G446 loop in BA.4 / 5 RBD are also caused by crystal contacts.(1) The 371-375 loop carrying S371F, S373P, and S375F mutations in BA.2.12.1 and BA.4 / 5 RBD is stabilized by interaction with CDR-H3 of NbCl. (m) Overlay of BA.2.12.1 (gray), BA.2 (green, PDB 7ZF9), and BA.4 / 5 (cyan) RBD. (n) Mutations at 452 do not introduce significant local structural changes. R452 in BA.4 / 5 has a dual conformation.

[0075] Figure 28 Vaccine neutralization assay against BA.4.6 pseudovirus. BA.1, BA.2, BA.4.5 immune sera (a-d) and monoclonal antibodies (e-f). IC50 values for the indicated viruses using sera obtained from vaccinees 28 days after the third dose of Pfizer BNT1622b2 vaccine (n=22, a). IC50 values for the indicated viruses against sera from volunteers who suffered vaccine breakthrough infection with BA.1 (n=14, b), BA.2 (n=23, c), and BA.4 / 5 (n=11, d). Geometric mean titers are shown above each column. Analysis was performed using Wilcoxon paired signed-rank test and two-tailed P values were calculated. Neutralization curves of a panel of 28 monoclonal antibodies made from samples collected from vaccinees infected with BA.1 against BA.4.6 (e) compared to Victoria, BA.1, BA.1.1, BA.2, BA.4 / 5, and BA.2.75 variants. Neutralization curves of a panel of 14 commercial monoclonal antibodies against the same variants (e). IC50 values are shown in Tables 29A and 29B.

[0076] Figure 29 Pseudovirus neutralization assay. Pseudovirus neutralization assay against Omicron monoclonal antibodies related to Table 26 where IC50 titers are shown. Neutralization curves of a panel of 27 monoclonal antibodies made from samples collected from vaccinees infected with BA.1. Titration curves of BA.2.11, BA.2.12.1, and BA.2.13 are compared to BA.2.

[0077] Figure 30 Surface plasmon resonance (SPR) analysis of the interaction between BA.2.12.1 or BA.2 RBD and selected mAbs (Omi-6 and Omi-31). (a) Affinity of BA.2.12.1 RBD for Omi-6 was determined using a 1:1 binding equilibrium analysis. (b), (c), (d) Sensograms (red: raw binding curves; black: fitted curves) showing the interaction between BA.2.12.1 or BA.2 RBD and selected mAbs, where kinetic data are shown.

[0078] Figure 31Neutralization assay. Neutralization curves obtained using lentiviruses pseudotyped with the S gene of the indicated BA.2 sublineage (A) Omi-mAbs, (B) commercial mAbs. See also Table 32. The “BA.4+all” variant is a synthetic variant designed after the evaluation of different mutations occurring in the SARS-CoV-2 Omicron S gene. These mutations were combined and incorporated into the Omicron BA.4 S gene to generate an artificial S gene called “BA.4+all”. This variant was generated only as an experimental tool, does not exist in nature and does not correspond to the S gene of any circulating SARS-Cov-2 variant.

[0079] Figure 32 Serum neutralization IC50 titers (dilution fold) of lentiviruses pseudotyped with the S gene of the indicated BA.2 sublineage. (A) sera obtained 28 days after the third dose of BNT162b2 vaccine or after infection with (B) BA.1, (C) BA.2 or (D) BA.4 / 5. Geometric mean titers are shown above each column. Wilcoxon matched-pairs signed-rank test (C and D) and Mann-Whitney test (E) were used and two-tailed P values were calculated.

[0080] Figure 33 Heatmap of antibody binding. Heatmap showing IC50 (pg / ml) of various antibodies against Victoria and Beta strains in vaccinated and unvaccinated samples.

[0081] Figure 34 Neutralization assay. Neutralization curves obtained using lentiviruses pseudotyped with the S gene of the indicated BA.2 sublineage.

[0082] Figure 35 Heatmap of IC50 neutralization titers of this panel of BA.1 (Omi) mAbs. The indicated mAbs were tested for pseudovirus neutralization IC50 titers against a panel of pseudoviruses expressing variant S sequences. Live virus IC50 values against variants found early in the pandemic are included for comparison. Live virus assay data and pseudovirus data for Victoria, BA.2 and BA.4 / 5 were previously reported in Tuekprakon et al. (2022). DETAILED DESCRIPTION

[0083] Antibodies of the invention

[0084] The antibodies of the application specifically bind to the spike protein of SARS-CoV-2.

[0085] In particular, it specifically binds to the S1 subunit of the spike protein, such as the receptor binding domain (RBD) or the N-terminal domain (NTD).

[0086] An antibody of the application can comprise at least three CDRs of an antibody in Table 3. The antibody can comprise at least four, five, or all six CDRs of an antibody in Table 3. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having at least 80% sequence identity to a heavy chain variable domain of an antibody in Table 3. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having at least 80% sequence identity to a light chain variable domain of an antibody in Table 3. The antibody can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having at least 80% identity to a heavy chain variable domain and a light chain variable domain, respectively, of an antibody in Table 3. The antibody can be any one of the antibodies in Table 3.

[0087] Table 3 lists 28 individual antibodies identified from breakthrough Omicron SARS-CoV-2 infected patients who have recovered and have received two doses of the Pfizer vaccine. Table 1 lists 42 individual antibodies previously identified from recovered COVID-19 patients [Dejnirattisai, Wanwisa, et al. “The antigenic anatomy of SARS-CoV-2 receptor binding domain.” Cell 184.8 (2021): 2183-2200; Supasa, Piyada, et al. “Reduced neutralization of SARS-CoV-2 B.1.1.7 variant by convalescent and vaccine sera.” Cell 184.8 (2021): 2201-2211; Zhou, Daming, et al. “Evidence of escape of SARS-CoV-2 variant B.1.351 from natural and vaccine-induced sera.” Cell 184.9 (2021): 2348-2361; Dejnirattisai, Wanwisa, et al. “Antibody evasion by the P.1 strain of SARS-CoV-2.” Cell 184.11 (2021): 2939-2954; Liu, Chang, et al. “Reduced neutralization of SARS-CoV-2 B.1.617 by vaccine and convalescent serum.” Cell 184.16 (2021): 4220-4236].Table 2 lists 28 individual antibodies previously identified from convalescent Beta SARS-CoV-2 infected patients [Liu, C, et al. “The antibody response to SARS-CoV-2 Beta underscores the antigenic distance to other variants”. Cell host & microbe 30(1) (2021): 53-68]. Antibodies in Table 1 are also referred to herein with the prefix “COVOX”, e.g. COVOX-222. Antibodies in Table 2 are also referred to with the prefix “Beta”, e.g. “Beta50”. Antibodies in Table 3 are also referred to with the prefix “O”, e.g. “O02”. Tables 1 to 3 list the SEQ ID NOs for the heavy chain variable region and light chain variable region nucleotide and amino acid sequences and the complementarity determining regions (CDRs) of the variable chains.

[0088] The antibodies in Table 3 can be selected from the group consisting of Omi03, Omi12, Omi02, Omi39, Omi42, Omi16, Omi18, Omi20, Omi 23, Omi28, Omi08, Omi17, Omi29, Omi36, and Omi38. Surprisingly, these antibodies were found to maintain strong neutralization against live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron (e.g. IC50 < 0.1 pg / ml against all live strains tested).

[0089] The antibodies in Table 3 can be selected from the group consisting of Omi03, Omi12, Omi02, Omi39, Omi42, Omi16, Omi18, Omi20, Omi 23, Omi28, and Omi 08. Surprisingly, these antibodies were found to maintain strong neutralization against live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron (e.g. IC50 < 0.05 pg / ml against all live strains tested).

[0090] The antibodies in Table 3 can be selected from the group consisting of Omi03, Omi12, Omi02, Omi39 and Omi42. Surprisingly, it was found that these antibodies maintain very strong neutralization against live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta and Omicron (e.g. IC50< 0.02 pg / ml against all live strains tested).

[0091] The antibodies in Table 3 can be selected from the group consisting of Omi03 and Omi12. Surprisingly, it was found that these antibodies maintain very strong neutralization against live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta and Omicron (e.g. IC50< 0.01 pg / ml against all live strains tested).

[0092] The antibodies in Table 3 can be selected from the group consisting of Omi02, Omi03, Omi12, Omi18, Omi28, Omi39 and Omi42. Surprisingly, it was found that these antibodies maintain very strong neutralization against live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta and Omicron.

[0093] Accordingly, in one embodiment, the antibody in Table 3 can be Omi03. It was found that Omi03 can neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In one embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3, which have the amino acid sequences set forth in SEQ ID NOs: 695, 696, and 697, respectively; and CDRL1, CDRL2, and CDRL3, which have the amino acid sequences set forth in SEQ ID NOs: 698, 699, and 700, respectively. In one embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi03 (i.e., SEQ ID NO: 692). In one embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi03 (i.e., SEQ ID NO: 694). In one embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, 100% sequence identity to the heavy chain variable domain and the light chain variable domain of antibody Omi03 (i.e., SEQ ID NOs: 692 and 694, respectively).

[0094] The heavy chain domain of Omi03 is derived from IGHV3-53 v-region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (explained further below). Thus, the antibodies of the present application can comprise the heavy chain of Omi03, but not the light chain of Omi03. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 695, 696 and 697, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi03 (i.e. SEQ ID NO: 692). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 692.

[0095] Alternatively, in one embodiment of the present application, the antibody can comprise the light chain of Omi03, but not the heavy chain of Omi03. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 698, 699 and 700, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Omi03 (i.e. SEQ ID NO: 694). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 694.

[0096] In an embodiment, the antibody in Table 3 can be Omi12. Omi12 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In an embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 735, 736, and 737, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 738, 739, and 740, respectively. In an embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi12 (i.e., SEQ ID NO: 732). In an embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi12 (i.e., SEQ ID NO: 734). In an embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and the light chain variable domain of antibody Omi12 (i.e., SEQ ID NOs: 732 and 734, respectively).

[0097] The heavy chain domain of Omi12 is derived from IGHV1-58 v-region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (explained further below). Thus, the antibodies of the present application can comprise the heavy chain of Omi12, but not the light chain of Omi12. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 735, 736 and 737, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi12 (i.e. SEQ ID NO: 732). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 732.

[0098] Alternatively, in one embodiment of the present application, the antibody can comprise the light chain of Omi12, but not the heavy chain of Omi12. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 738, 739 and 740, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Omi12 (i.e. SEQ ID NO: 734). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 734.

[0099] In an embodiment, the antibody in Table 3 can be Omi02. Omi02 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In an embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 685, 686, and 687, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 688, 689, and 690, respectively. In an embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi02 (i.e., SEQ ID NO: 682). In an embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi02 (i.e., SEQ ID NO: 684). In an embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and the light chain variable domain of antibody Omi02 (i.e., SEQ ID NOs: 682 and 684, respectively).

[0100] The heavy chain domain of Omi02 is derived from the IGHV1-69 v-region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (explained further below). Thus, the antibodies of the present application can comprise the heavy chain of Omi02, but not the light chain of Omi02. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 685, 686 and 687, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi02 (i.e. SEQ ID NO: 682). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 682.

[0101] Alternatively, in one embodiment of the present application, the antibody can comprise the light chain of Omi02, but not the heavy chain of Omi02. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 688, 689 and 690, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Omi02 (i.e. SEQ ID NO: 684). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 684.

[0102] In an embodiment, the antibody in Table 3 can be Omi08. It was found that Omi08 can neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In an embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 715, 716, and 717, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 718, 719, and 720, respectively. In an embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi08 (i.e., SEQ ID NO: 712). In an embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi08 (i.e., SEQ ID NO: 714). In an embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, 100% sequence identity to the heavy chain variable domain and the light chain variable domain of antibody Omi08 (i.e., SEQ ID NOs: 712 and 714, respectively).

[0103] In an embodiment, the antibody in Table 3 can be Omi42. Omi42 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In an embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 955, 956, and 957, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 958, 959, and 960, respectively. In an embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi42 (i.e., SEQ ID NO: 952). In an embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi42 (i.e., SEQ ID NO: 954). In an embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and the light chain variable domain of antibody Omi42 (i.e., SEQ ID NOs: 952 and 954, respectively).

[0104] The heavy chain domain of Omi42 is derived from the IGHV3-9 v-region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (explained further below). Thus, the antibodies of the present application can comprise the heavy chain of Omi42, but not the light chain of Omi42. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 955, 956 and 957, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi42 (i.e. SEQ ID NO: 952). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 952.

[0105] Alternatively, in one embodiment of the present application, the antibody can comprise the light chain of Omi42, but not the heavy chain of Omi42. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 958, 959 and 960, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Omi42 (i.e. SEQ ID NO: 954). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 954.

[0106] In one embodiment, the antibody in Table 3 can be Omi16. Omi16 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In one embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3, having the amino acid sequences set forth in SEQ ID NOs: 745, 746, and 747, respectively; and CDRL1, CDRL2, and CDRL3, having the amino acid sequences set forth in SEQ ID NOs: 748, 749, and 750, respectively. In one embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi16 (i.e., SEQ ID NO: 742). In one embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi16 (i.e., SEQ ID NO: 744). In one embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and light chain variable domain of antibody Omi16 (i.e., SEQ ID NOs: 742 and 744, respectively).

[0107] The heavy chain domain of Omi16 is derived from the IGHV3-66 v-region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (explained further below). Thus, the antibodies of the present application can comprise the heavy chain of Omi16, but not the light chain of Omi16. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences specified in SEQ ID NOs: 745, 746 and 747, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi16 (i.e. SEQ ID NO: 742). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 742.

[0108] Alternatively, in one embodiment of the application, the antibody can comprise the light chain of Omi16, but not the heavy chain of Omi16. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences specified in SEQ ID NOs: 748, 749 and 750, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Omi16 (i.e. SEQ ID NO: 744). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 744.

[0109] In an embodiment, the antibody in Table 3 can be Omi18. Omi18 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In an embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 765, 766, and 767, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 768, 769, and 770, respectively. In an embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi18 (i.e., SEQ ID NO: 762). In an embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi18 (i.e., SEQ ID NO: 764). In an embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and light chain variable domain of antibody Omi18 (i.e., SEQ ID NOs: 762 and 764, respectively).

[0110] The heavy chain domain of Omi18 is derived from the IGHV3-53 v-region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (explained further below). Thus, the antibody of the present application can comprise the heavy chain of Omi18, but not the light chain of Omi18. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 765, 766 and 767, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi18 (i.e. SEQ ID NO: 762).

[0111] The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 762.

[0112] Alternatively, in one embodiment of the present application, the antibody can comprise the light chain of Omi18, but not the heavy chain of Omi18. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 768, 769 and 770, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Omi18 (i.e. SEQ ID NO: 764). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 764.

[0113] In an embodiment, the antibody in Table 3 can be Omi20. Omi20 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In an embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 775, 776, and 777, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 778, 779, and 780, respectively. In an embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi20 (i.e., SEQ ID NO: 772). In an embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi20 (i.e., SEQ ID NO: 774). In an embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and the light chain variable domain of antibody Omi20 (i.e., SEQ ID NOs: 772 and 774, respectively).

[0114] The heavy chain domain of Omi20 is derived from the IGHV3-66 v-region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (explained further below). Thus, the antibodies of the present application can comprise the heavy chain of Omi20, but not the light chain of Omi20. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 775, 776 and 777, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi20 (i.e. SEQ ID NO: 772). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 772.

[0115] Alternatively, in one embodiment of the present application, the antibody can comprise the light chain of Omi20, but not the heavy chain of Omi20. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 778, 779 and 780, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Omi20 (i.e. SEQ ID NO: 774). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 774.

[0116] In an embodiment, the antibody in Table 3 can be Omi23. Omi23 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In an embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 785, 786, and 787, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 788, 789, and 790, respectively. In an embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi23 (i.e., SEQ ID NO: 782). In an embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi23 (i.e., SEQ ID NO: 784). In an embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and the light chain variable domain of antibody Omi23 (i.e., SEQ ID NOs: 782 and 784, respectively).

[0117] The heavy chain domain of Omi23 is derived from the IGHV4-31 v-region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (explained further below). Thus, the antibodies of the present application can comprise the heavy chain of Omi23, but not the light chain of Omi23. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 785, 786 and 787, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi23 (i.e. SEQ ID NO: 782). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 782.

[0118] Alternatively, in one embodiment of the present application, the antibody can comprise the light chain of Omi23, but not the heavy chain of Omi23. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 788, 789 and 790, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Omi23 (i.e. SEQ ID NO: 784). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 784.

[0119] In an embodiment, the antibody in Table 3 can be Omi28. Omi28 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In an embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 835, 836, and 837, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 838, 839, and 840, respectively. In an embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi28 (i.e., SEQ ID NO: 832). In an embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi28 (i.e., SEQ ID NO: 834). In an embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and light chain variable domain of antibody Omi28 (i.e., SEQ ID NOs: 832 and 834, respectively). The heavy chain domain of Omi28 is derived from the IGHV3-66v- region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (further explained below). Thus, an antibody of the application can comprise the heavy chain of Omi28, but not the light chain of Omi28. For example, the antibody can comprise CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 835, 836, and 837, respectively.The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity with the heavy chain variable domain of antibody Omi28 (i.e., SEQ ID NO: 832). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 832.

[0120] Alternatively, in one embodiment of the application, the antibody can comprise the light chain of Omi28, but not the heavy chain of Omi28. For example, the antibody can comprise CDRL1, CDRL2, and CDRL3, each having the amino acid sequence specified in SEQ ID NO: 838, 839, and 840, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity with the light chain variable domain of antibody Omi28 (i.e., SEQ ID NO: 834). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 834.

[0121] In an embodiment, the antibody in Table 3 can be Omi39. Omi39 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In an embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 935, 936, and 937, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 938, 939, and 940, respectively. In an embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi39 (i.e., SEQ ID NO: 932). In an embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi39 (i.e., SEQ ID NO: 934). In an embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and light chain variable domain of antibody Omi39 (i.e., SEQ ID NOs: 932 and 934, respectively).

[0122] In one embodiment, the antibody in Table 3 can be Omi17. Omi17 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In one embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3, having the amino acid sequences set forth in SEQ ID NOs: 755, 756, and 757, respectively; and CDRL1, CDRL2, and CDRL3, having the amino acid sequences set forth in SEQ ID NOs: 758, 759, and 760, respectively. In one embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi17 (i.e., SEQ ID NO: 752). In one embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi17 (i.e., SEQ ID NO: 754). In one embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and light chain variable domain of antibody Omi17 (i.e., SEQ ID NOs: 752 and 754, respectively).

[0123] The heavy chain domain of Omi17 is derived from the IGHV3-66 v-region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (explained further below). Thus, the antibodies of the present application can comprise the heavy chain of Omi17, but not the light chain of Omi17. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences specified in SEQ ID NOs: 755, 756 and 757, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi17 (i.e. SEQ ID NO: 752). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 752.

[0124] Alternatively, in one embodiment of the application, the antibody can comprise the light chain of Omi17, but not the heavy chain of Omi17. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences specified in SEQ ID NOs: 758, 759 and 760, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Omi17 (i.e. SEQ ID NO: 754). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 754.

[0125] In one embodiment, the antibody in Table 3 can be Omi29. Omi29 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In one embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 845, 846, and 847, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 848, 849, and 850, respectively. In one embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi29 (i.e., SEQ ID NO: 842). In one embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi29 (i.e., SEQ ID NO: 844). In one embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and the light chain variable domain of antibody Omi29 (i.e., SEQ ID NOs: 842 and 844, respectively).

[0126] The heavy chain domain of Omi29 is derived from the IGHV3-53 v-region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (explained further below). Thus, the antibodies of the present application can comprise the heavy chain of Omi29, but not the light chain of Omi29. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 845, 846 and 847, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi29 (i.e. SEQ ID NO: 842). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 842.

[0127] Alternatively, in one embodiment of the present application, the antibody can comprise the light chain of Omi29, but not the heavy chain of Omi29. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 848, 849 and 850, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Omi29 (i.e. SEQ ID NO: 844). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 844.

[0128] In an embodiment, the antibody in Table 3 can be Omi36. Omi36 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In an embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 915, 916, and 917, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 918, 919, and 920, respectively. In an embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi36 (i.e., SEQ ID NO: 912). In an embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi36 (i.e., SEQ ID NO: 914).

[0129] In one embodiment, the antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain and light chain variable domain of antibody Omi36 (i.e. SEQ ID NOs: 912 and 914, respectively). The heavy chain domain of Omi36 is derived from the IGHV3-66 v-region, and the inventors have previously demonstrated that switching heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for use in the application (explained further below). Thus, the antibody of the application can comprise the heavy chain of Omi36, but not the light chain of Omi36. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences specified in SEQ ID NOs: 915, 916 and 917, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi36 (i.e. SEQ ID NO: 912). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 912.

[0130] Alternatively, in one embodiment of the application, the antibody can comprise the light chain of Omi36, but not the heavy chain of Omi36. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences specified in SEQ ID NOs: 918, 919 and 920, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Omi36 (i.e. SEQ ID NO: 914). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 914.

[0131] In an embodiment, the antibody in Table 3 can be Omi38. Omi38 was found to neutralize live SARS-CoV-2 variant strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron, as well as pseudovirus constructs of Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3 (see Tables 13 and 14, Figure 2). In an embodiment, an antibody of the application can comprise: CDRH1, CDRH2, and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 925, 926, and 927, respectively; and CDRL1, CDRL2, and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 928, 929, and 930, respectively. In an embodiment, an antibody of the application can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain of antibody Omi38 (i.e., SEQ ID NO: 922). In an embodiment, an antibody of the application can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of antibody Omi38 (i.e., SEQ ID NO: 924). In an embodiment, an antibody of the application can comprise a heavy chain variable domain and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the heavy chain variable domain and the light chain variable domain of antibody Omi38 (i.e., SEQ ID NOs: 922 and 924, respectively).

[0132] The heavy chain domain of Omi38 is derived from the IGHV1-69 v-region, and the inventors have previously demonstrated that switching the heavy and light chains between antibodies derived from the same v-region can yield antibodies particularly suitable for the present application (explained further below). Thus, the antibodies of the present application can comprise the heavy chain of Omi38, but not the light chain of Omi38. For example, the antibody can comprise CDRH1, CDRH2 and CDRH3 having the amino acid sequences set forth in SEQ ID NOs: 925, 926 and 927, respectively. The antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the heavy chain variable domain of antibody Omi38 (i.e., SEQ ID NO: 922). The antibody can comprise a heavy chain variable domain comprising or consisting of SEQ ID NO: 922.

[0133] Alternatively, in one embodiment of the application, the antibody can comprise the light chain of Omi38, but not the heavy chain of Omi38. For example, the antibody can comprise CDRL1, CDRL2 and CDRL3 having the amino acid sequences set forth in SEQ ID NOs: 928, 929 and 930, respectively. The antibody can comprise a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99% or 100% sequence identity to the light chain variable domain of antibody Om38 (i.e., SEQ ID NO: 924). The antibody can comprise a light chain variable domain comprising or consisting of SEQ ID NO: 924.

[0134] Mixed chain antibodies of the invention

[0135] The antibodies of the present application can comprise a light chain variable domain comprising CDRL1, CDRL2 and CDRL3 from a first antibody in Tables 1, 2 or 3, and a heavy chain variable domain comprising CDRH1, CDRH2 and CDRH3 from a second antibody in Tables 1, 2 or 3, provided that the first and second antibodies are different. Such antibodies are referred to herein as mixed chain antibodies.

[0136] Examples of hybrid chain antibodies useful in the present application are provided in Tables 4-12. Table 4 shows examples of hybrid chain antibodies derived from the same germline heavy chain IGHV 3-53 produced from the antibodies in Tables 1-3. Table 5 shows examples of hybrid chain antibodies derived from the same germline heavy chain IGHV 3-53 and IGHV 3-66 produced from the antibodies in Tables 1-3. Table 6 shows examples of hybrid chain antibodies derived from the same germline heavy chain IGHV 1-58 produced from the antibodies in Tables 1-3. Table 7 shows examples of hybrid chain antibodies derived from the same germline heavy chain IGHV 1-69 produced from the antibodies in Tables 2 and 3. Table 8 shows examples of hybrid chain antibodies derived from the same germline heavy chain IGHV 3-30 produced from the antibodies in Tables 1-3. Table 9 shows examples of hybrid chain antibodies derived from the same germline heavy chain IGHV 3-33 produced from the antibodies in Tables 2 and 3. Table 10 shows examples of hybrid chain antibodies derived from the same germline heavy chain IGHV 1-18 produced from the antibodies in Tables 1-3. Table 11 shows examples of hybrid chain antibodies derived from the same germline heavy chain IGHV 3-9 produced from the antibodies in Tables 1 and 3. Table 12 shows examples of hybrid chain antibodies derived from the same germline heavy chain IGHV 4-31 produced from the antibodies in Tables 2 and 3. Examples of hybrid chain antibodies derived from the same germline heavy chain IGHV 1-69 are Omi02H / Beta-49L and Omi38H / Omi24L.

[0137] Accordingly, in one embodiment, an antibody of the application comprises a heavy chain variable domain comprising CDRH1, CDRH2, and CDRH3 from a first antibody in Table 1, 2, or 3, and a light chain variable domain comprising CDRL1, CDRL2, and CDRL3 from a second antibody in Table 1, 2, or 3, provided that the first antibody and the second antibody are different. The antibody can comprise a heavy chain variable domain amino acid sequence having at least 80% sequence identity to a heavy chain variable domain from a first antibody in Table 1, 2, or 3, and a light chain variable domain amino acid sequence having at least 80% sequence identity to a light chain variable domain from a second antibody in Table 1, 2, or 3, provided that the first antibody and the second antibody are different. For example, the antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to a heavy chain variable domain of an antibody in Table 1, 2, or 3, and a light chain variable domain comprising or consisting of an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to a light chain variable domain of an antibody in Table 1, 2, or 3, provided that the first antibody and the second antibody are different.

[0138] The first antibody can be in Table 3, and the second antibody can be in Table 3.

[0139] The first antibody can be in Table 3, and the second antibody can be in Table 1. The first antibody can be in Table 3, and the second antibody can be in Table 2. The first antibody can be in Table 1, and the second antibody can be in Table 3. The first antibody can be in Table 2, and the second antibody can be in Table 3. The first antibody can be in Table 1, and the second antibody can be in Table 2. The first antibody can be in Table 2, and the second antibody can be in Table 1. The first antibody can be in Table 2, and the second antibody can be in Table 2. The first antibody can be in Table 1, and the second antibody can be in Table 1.

[0140] In one embodiment, at least one of the first antibody and the second antibody is an antibody from Table 3.

[0141] In one embodiment, the first antibody and the second antibody are not both in Table 1. In one embodiment, the first antibody and the second antibody are not both in Table 2. In one embodiment, the first antibody and the second antibody are not both selected from the antibodies in Table 1 or 2.

[0142] In one embodiment, at least one of the heavy chain variable domain and the light chain variable domain is from Table 3.

[0143] The antibody in Table 3 can be selected from the group consisting of Omi02, Omi03, Omi12, Omi18, Omi28, Omi39, and Omi42. The antibody in Table 3 can be selected from the group consisting of Omi03, Omi12, Omi02, Omi39, Omi42, Omi16, Omi18, Omi20, Omi23, Omi28, Omi08, Omi17, Omi29, Omi36, and Omi38. For example, the antibody in Table 3 can be selected from the group consisting of Omi03, Omi12, Omi02, Omi39, Omi42, Omi16, Omi18, Omi20, Omi23, Omi28, and Omi08. The antibody in Table 3 can be selected from the group consisting of Omi03, Omi12, Omi02, Omi39, and Omi42. The antibody in Table 3 can be selected from the group consisting of Omi03 and Omi12.

[0144] In one embodiment, both the first antibody and the second antibody are selected from the group consisting of Omi03, Omi18, Omi29, Beta-27, antibody 150, antibody 158, antibody 175, antibody 222, and antibody 269. The heavy chain variable domains of these antibodies are derived from IGHV3-53. The resulting hybrid chain antibodies are listed in Table 4. Accordingly, the antibodies of the present application can comprise all six CDRs (CDRH1-3 and CDRL1-3), and / or a heavy chain variable domain and a light chain variable domain each comprising or consisting of an amino acid sequence having at least 80% sequence identity to the corresponding variable domain of any one of the hybrid chain antibodies as listed in Table 4.

[0145] Antibodies derived from IGHV3-53 can be used to create a hybrid chain antibody with an antibody from IGHV3-66 (e.g., antibodies 40 and 398 in Table 1) (see, e.g., Dejnirattisai, Wanwisa, et al. “The antigenic anatomy of SARS-CoV-2 receptor binding domain.” Cell 184.8 (2021): 2183-2200; Supasa, Piyada, et al. “Reduced neutralization of SARS-CoV-2 B.1.1.7 variant by convalescent and vaccine sera.” Cell 184.8 (2021): 2201-2211; Zhou, Daming, et al. “Evidence of escape of SARS-CoV-2 variant B.1.351 from natural and vaccine-induced sera.” Cell 184.9 (2021): 2348-2361; Dejnirattisai, Wanwisa, et al. “Antibody evasion by the P.1 strain of SARS-CoV-2.” Cell 184.11 (2021): 2939-2954; Liu, Chang, et al. “Reduced neutralization of SARS-CoV-2 B.1.617 by vaccine and convalescent serum.” Cell 184.16 (2021): 4220-4236). Thus, in one embodiment, the first antibody and the second antibody are both selected from the group consisting of Omi03, Omi18, Omi29, Omi16, Omi17, Omi20, Omi27, Omi36, Beta-27, antibody 150, antibody 158, antibody 175, antibody 222, antibody 269, antibody 40, and antibody 398. The heavy chain variable domains of these antibodies are derived from IGHV3-53 and IGVH3-66.The resulting hybrid chain antibodies are listed in Table 5. Thus, antibodies of the application can comprise all six CDRs (CDRH1-3 and CDRL1-3), and / or a heavy chain variable domain and a light chain variable domain each comprising or consisting of an amino acid sequence having at least 80% (e.g., >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100%) sequence identity to the corresponding variable domain of any one of the hybrid chain antibodies as listed in Table 5.

[0146] In one embodiment, the first antibody and the second antibody are both selected from the group consisting of Omi12, Beta-47, Beta-25, Antibody 55, Antibody 165, Antibody 253, and Antibody 318. The heavy chain variable domains of these antibodies are derived from IGHV1-58. The resulting hybrid chain antibodies are listed in Table 6. Thus, antibodies of the application can comprise all six CDRs (CDRH1-3 and CDRL1-3), and / or a heavy chain variable domain and a light chain variable domain each comprising or consisting of an amino acid sequence having at least 80% (e.g., >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100%) sequence identity to the corresponding variable domain of any one of the hybrid chain antibodies as listed in Table 6.

[0147] In one embodiment, the first antibody and the second antibody are both selected from the group consisting of Beta-49, Beta-50, Omi02, Omi24, Omi30, Omi31, Omi34, and Omi38. The heavy chain variable domains of these antibodies are derived from IGHV 1-69. The resulting hybrid chain antibodies are listed in Table 7. Thus, antibodies of the application can comprise all six CDRs (CDRH1-3 and CDRL1-3), and / or a heavy chain variable domain and a light chain variable domain each comprising or consisting of an amino acid sequence having at least 80% (e.g., >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100%) sequence identity to the corresponding variable domain of any one of the hybrid chain antibodies as listed in Table 7.

[0148] In one embodiment, the first antibody and the second antibody are both selected from the group consisting of Beta-22, Beta-29, Antibody 159, and Omi09. The heavy chain variable domains of these antibodies are derived from IGHV 3-30. The resulting hybrid chain antibodies are listed in Table 8. Thus, an antibody of the present application can comprise all six CDRs (CDRH1-3 and CDRL1-3), and / or a heavy chain variable domain and a light chain variable domain each comprising or consisting of an amino acid sequence having at least 80% (e.g., >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100%) sequence identity to the corresponding variable domain of any one of the hybrid chain antibodies as listed in Table 8.

[0149] In one embodiment, the first antibody and the second antibody are both selected from the group consisting of Beta-20, Beta-43, Omi32, and Omi33. The heavy chain variable domains of these antibodies are derived from IGHV 3-33. The resulting hybrid chain antibodies are listed in Table 9. Thus, an antibody of the present application can comprise all six CDRs (CDRH1-3 and CDRL1-3), and / or a heavy chain variable domain and a light chain variable domain each comprising or consisting of an amino acid sequence having at least 80% (e.g., >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100%) sequence identity to the corresponding variable domain of any one of the hybrid chain antibodies as listed in Table 9. The CDRL1-3 of Omi32 and Omi33 are identical, meaning that they are in fact already exemplary hybrid chain antibodies of the present application.

[0150] In one embodiment, the first antibody and the second antibody are both selected from the group consisting of Antibody 278, Beta-44, Omi26, and Omi41. The heavy chain variable domains of these antibodies are derived from IGHV 1-18. The resulting hybrid chain antibodies are listed in Table 10. Thus, an antibody of the present application can comprise all six CDRs (CDRH1-3 and CDRL1-3), and / or a heavy chain variable domain and a light chain variable domain each comprising or consisting of an amino acid sequence having at least 80% (e.g., >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100%) sequence identity to the corresponding variable domain of any one of the hybrid chain antibodies as listed in Table 10.

[0151] In one embodiment, the first antibody and the second antibody are both selected from the group consisting of antibody 58, Omi25, Omi35, and Omi42. The heavy chain variable domains of these antibodies are derived from IGHV 3-9. The resulting hybrid chain antibodies are listed in Table 11. Thus, an antibody of the application can comprise all six CDRs (CDRH1-3 and CDRL1-3), and / or a heavy chain variable domain and a light chain variable domain each comprising or consisting of an amino acid sequence having at least 80% (e.g., >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100%) sequence identity to the corresponding variable domain of any one of the hybrid chain antibodies as listed in Table 11.

[0152] In one embodiment, the first antibody and the second antibody are both selected from the group consisting of Beta-56 and Omi23. The heavy chain variable domains of these antibodies are derived from IGHV 4-31. The resulting hybrid chain antibodies are listed in Table 12. Thus, an antibody of the application can comprise all six CDRs (CDRH1-3 and CDRL1-3), and / or a heavy chain variable domain and a light chain variable domain each comprising or consisting of an amino acid sequence having at least 80% (e.g., >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100%) sequence identity to the corresponding variable domain of any one of the hybrid chain antibodies as listed in Table 12.

[0153] The constant region domain of an antibody molecule of the application (if present) can be selected with consideration of the intended function of the antibody molecule, in particular the effector functions that can be desired. For example, the constant region domain can be an IgA, IgD, IgE, IgG, or IgM domain of human origin. Typically, the constant region is of human origin. In particular, a human IgG (i.e., IgGl, IgG2, IgG3, or IgG4) constant region domain can be used. Typically, the constant region is a human IgGl constant region.

[0154] Certain antibodies of the invention

[0155] The application also provides an antibody that is a full-length antibody of any one of the antibodies in Tables 3, 4, 5, 6, 7, 8, 9, 10, 11, or 12. In other words, an antibody of the application comprises a heavy chain variable domain and a light chain variable domain that each consist of the heavy chain variable domain and the light chain variable domain, respectively, of any one of the antibodies in Tables 3, 4, 5, 6, 7, 8, 9, 10, 11, or 12, and an IgG (e.g., IgGl) constant region.

[0156] For example, an antibody of the application can be a full-length Omi02, Omi03, Omi12, Omi18, Omi28, Omi39, or Omi42 antibody. An antibody of the application can be a full-length Omi03, Omi12, Omi02, Omi39, Omi42, Omi16, Omi18, Omi20, Omi23, Omi28, Omi08, Omi17, Omi29, Omi36, or Omi38 antibody. These antibodies are all highly potent neutralizing mAbs that have been shown to neutralize the Omicron variant of SARS-CoV-2 with IC50≥ 0.1 μg / ml. These antibodies also specifically remain neutralizing against at least Victoria, Alpha, Beta, Gamma, and Delta strains of SARS-CoV-2 with IC50≤ 0.1 μg / ml.

[0157] The antibody can be derived from the germline heavy chain IGHV1-58 and comprise a proline at position 53 in the heavy chain variable region (according to the absolute numbering). For example, the antibody can comprise a heavy chain variable domain comprising or consisting of an amino acid sequence having ≥ 80%, ≥ 90%, ≥ 95%, ≥ 96%, ≥ 97%, ≥ 98%, ≥ 99%, or 100% sequence identity with the heavy chain variable domain of Omi-12 (SEQ ID NO: 731), Beta-47 (SEQ ID NO: 591), Beta-25 (SEQ ID NO: 461), antibody 55 (SEQ ID NO: 62), antibody 165 (SEQ ID NO: 182), antibody 253 (SEQ ID NO: 262), or antibody 318 (SEQ ID NO: 332), provided that the amino acid at position 53 in the heavy chain variable region is a proline (according to the absolute numbering). For example, the antibody can comprise the heavy chain variable region and the light chain variable region of Beta-47 (SEQ ID NO: 591 and 592, respectively), Beta-25 (SEQ ID NO: 461 and 462, respectively), antibody 55 (SEQ ID NO: 62 and 61, respectively), antibody 165 (SEQ ID NO: 182 and 181, respectively), antibody 253 (SEQ ID NO: 262 and 261, respectively), or antibody 318 (SEQ ID NO: 332 and 331, respectively), except for the V53P mutation in the heavy chain variable region. The inventors found that such antibodies are particularly potent against the Omicron strain (see, e.g., Example 5).

[0158] Position 53 of the heavy chain variable region of the IGHV1-58-derived antibodies Omi-12, Beta-47, Beta-25, antibody 55, antibody 165, antibody 253, and antibody 318 corresponds to position 58 according to IMGT numbering.

[0159] Accordingly, the present application also provides an antibody derived from the germline heavy chain IGHV1-58, which is capable of binding to the spike protein of the coronavirus SARS-CoV-2, wherein the amino acid at position 58 according to IMGT numbering in the heavy chain variable region is proline or is substituted by proline.

[0160] The antibody can comprise a heavy chain variable domain comprising an amino acid sequence having > 80%, > 90%, > 95%, > 96%, > 97%, > 98%, > 99% or 100% sequence identity to the heavy chain variable domain of an antibody derived from the germline heavy chain IGHV1-58, with the proviso that the amino acid at position 58 according to IMGT numbering is proline or is substituted by proline.

[0161] The antibody derived from the germline heavy chain IGHV1-58 can be AZD8895, Omi-12, Beta-47, Beta-25, antibody 55, antibody 165, antibody 253 or antibody 318. The amino acid sequences of the heavy chain variable domains of Omi-12, Beta-47, Beta-25, antibody 55, antibody 165, antibody 253 or antibody 318 are described herein (see, e.g., Tables 1 to 3). The amino acid sequence of the heavy chain variable domain of antibody AZD8895 is provided in SEQ ID NO: 963.

[0162] The IGHV1-58 germline V gene sequence encodes the following amino acid sequence: MQLVQSGPEVKKPGTSVKVSCKASGFTFTSSAVQWVRQARGQRL EWIGWIVVGSGNTNYAQKFQERVTITRDMSTSTAYMELSSLRSED TAVYYCAA (SEQ ID NO: 961). Accordingly, the present application also provides an antibody capable of binding to the spike protein of the coronavirus SARS-CoV-2, which comprises a heavy chain variable domain comprising an amino acid sequence having > 60%, > 70%, > 80%, > 90%, > 95%, > 96%, > 97%, > 98%, > 99% or 100% sequence identity to SEQ ID NO: 961, with the proviso that the amino acid at position 58 according to IMGT numbering is proline or is substituted by proline.

[0163] The antibody can comprise a heavy chain variable domain comprising an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to SEQ ID NO: 731, 591, 461, 62, 182, 262, 332, or 963, provided that the amino acid at position 58 according to IMGT numbering is proline or is substituted with proline. The antibody can comprise a heavy chain variable domain comprising an amino acid sequence having SEQ ID NO: 591, 461, 62, 182, 262, or 332, wherein the valine at position 58 according to IMGT numbering is substituted with proline.

[0164] The antibody can comprise a heavy chain variable domain comprising an amino acid sequence having SEQ ID NO: 963, wherein the isoleucine at position 58 according to IMGT numbering is substituted with proline.

[0165] In some embodiments, the antibody derived from the germline heavy chain IGHV1-58 comprises a light chain variable domain derived from IGLV Kappa 3-20. The antibody can comprise a light chain variable domain comprising an amino acid sequence having >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to the light chain variable domain of an antibody derived from the germline IGLV Kappa 3-20. The germline IGLV Kappa 3-20 V sequence can encode the following amino acid sequence: EIVLTQSPGTLSLSPGERATLSCRASQSV SSSYLAWYQQKPGQAPRLLIYGASSRATGIPDRFSGSGSGTDFTLTI SRLEPEDFAVYYCQQYGSSP (SEQ ID NO: 967). Thus, the antibody derived from the germline heavy chain IGHV1-58 can comprise a light chain variable domain comprising an amino acid sequence having >60%, >70%, >80%, >90%, >95%, >96%, >97%, >98%, >99%, or 100% sequence identity to SEQ ID NO: 967.

[0166] The present application also provides an antibody capable of binding to the spike protein of the coronavirus SARS-CoV-2, the antibody comprising a heavy chain variable domain comprising an amino acid sequence which is a modified version of SEQ ID NO: 961, provided that the amino acid at position 58 according to IMGT numbering is proline or is substituted by proline. The modified version of SEQ ID NO: 961 can comprise modifications as described herein, e.g. substitutions, deletions and / or additions. For example, the modification can comprise < 50, < 45, < 40, < 35, < 30, < 25, < 20, < 15, < 10, < 9, < 8, < 7, < 6, < 5, < 4, < 3, < 2, or 1 amino acid substitution(s) and / or deletion(s) of SEQ ID NO: 961. The modification can comprise < 4, < 3, < 2, or 1 amino acid substitution(s) and / or deletion(s) of SEQ ID NO: 961.

[0167] The antibody can comprise a heavy chain variable domain comprising an amino acid sequence which is a modified version of SEQ ID NO: 731, 591, 461, 62, 182, 262, 332, or 963, the modified version comprising < 10, < 9, < 8, < 7, < 6, < 5, < 4, < 3, < 2, or 1 modification, provided that the amino acid at position 58 according to IMGT numbering is proline or is substituted by proline. The modified version of SEQ ID NO: 731, 591, 461, 62, 182, 262, 332, or 963 can comprise modifications as described herein, e.g. substitutions, deletions and / or additions.

[0168] The antibody can comprise an IgG (e.g. IgG1) constant region.

[0169] The present application also provides a method of making such an antibody. For example, the method can comprise modifying an antibody derived from the germline heavy chain IGHV1-58, the antibody being capable of binding to the spike protein of the coronavirus SARS-CoV-2, by substituting the amino acid at position 58 (according to IMGT numbering) in the heavy chain variable region with proline. The antibody derived from the germline heavy chain IGHV1-58 can be AZD8895, Omi-12, Beta-47, Beta-25, Antibody 55, Antibody 165, Antibody 253, or Antibody 318. The amino acid sequence of the heavy chain variable domain of each of these antibodies is described herein (see, e.g., Tables 1 to 3 and SEQ ID NO: 963). The present application also provides an antibody obtainable or obtained by the method.

[0170] Properties of antibodies of the invention

[0171] The antibodies of the present application can be or can comprise modifications of the amino acid sequences from the antibodies in Tables 1-12 while maintaining the activity and / or function of the antibody. The modifications can be substitutions, deletions, and / or additions. For example, the modifications can include 1, 2, 3, 4, 5, up to 10, up to 20, up to 30, or more amino acid substitutions and / or deletions from the amino acid sequences of the antibodies in Tables 1-12. For example, the modifications can include substitution of an amino acid with an alternative amino acid having similar properties. Some properties of the 20 common amino acids that can be used to select a suitable substitution are as follows:

[0172] Ala Aliphatic, hydrophobic, neutral Met Hydrophobic, neutral Cys Polar, hydrophobic, neutral Asn Polar, hydrophilic, neutral Asp Polar, hydrophilic, charged (-) Pro Hydrophobic, neutral Glu Polar, hydrophilic, charged (-) Gln Polar, hydrophilic, neutral Phe Aromatic, hydrophobic, neutral Arg Polar, hydrophilic, charged (+) Gly Aliphatic, neutral Ser Polar, hydrophilic, neutral His Aromatic, polar, hydrophilic, charged (+) Thr Polar, hydrophilic, neutral Ile Aliphatic, hydrophobic, neutral Val Aliphatic, hydrophobic, neutral Lys Polar, hydrophilic, charged (+) Trp Aromatic, hydrophobic, neutral Leu Aliphatic, hydrophobic, neutral Tyr Aromatic, polar, hydrophobic

[0173] The modifications can include derivatized amino acids, such as labeled or non-natural amino acids, provided that the function of the antibody is not significantly adversely affected.

[0174] Modifications of the antibodies of the present application as described above can be made during the synthesis of the antibody or by post-production modification, or when the antibody is in recombinant form, using known site-directed mutagenesis, random mutagenesis, or enzymatic cleavage and / or ligation of nucleic acids techniques.

[0175] The antibodies of the present application can be modified (e.g., as described above) to improve the potency of the antibody or to adapt the antibody to new SARS-CoV-2 variants. These modifications can be amino acid substitutions to adapt the antibody to substitutions in the virus variants. For example, the known binding mode of the antibody to the spike protein (e.g., by crystal structure determination or modeling) can be used to identify amino acids of the antibody that interact with substitutions in the virus variants. This information can then be used to identify possible substitutions of the antibody that will compensate for changes in the epitope characteristics. For example, substitution of a hydrophobic amino acid in the spike protein to a negatively charged amino acid can be compensated for by substitution of an amino acid from the antibody that interacts with the amino acid in the spike protein to a positively charged amino acid. The present disclosure includes methods for identifying residues of the antibody that can be substituted, e.g., by determining the structure of the antibody-antigen complex as described herein.

[0176] The antibodies of the present application can contain one or more modifications to increase their cross-lineage neutralization properties. For example, the E484 of the spike protein, which is a key residue that mediates interaction with ACE2, is mutated in some SARS-CoV-2 strains (e.g., Victoria strain contains E484, but P.1 and B.1.351 strains contain E484K), resulting in different neutralization effects of antibodies. Thus, antibodies that bind to E484 can be modified to compensate for the change in E484 of the spike protein. For example, E484 mutates from a positively charged amino acid to a negatively charged amino acid in SARS-CoV-2 strains of the B.1.351 or P.1 lineage when compared to the original strain. Amino acid residues of the antibody that bind or are close to E484 can be mutated to compensate for the change in charge. Examples of such amino acid residues can be G104 and / or K108 in SEQ ID NO: 102 of antibody 88 or R52 in SEQ ID NO: 372 of antibody 384.

[0177] The antibodies of the present application can be isolated antibodies. An isolated antibody is an antibody that is substantially free of other antibodies having different antigenic specificities.

[0178] The term "antibody" as used herein can relate to intact antibodies (i.e., comprising two heavy chains and two light chains interconnected by disulfide bonds) as well as antigen-binding fragments thereof. Antibodies generally comprise an immunologically active portion of an immunoglobulin (Ig) molecule (i.e., a molecule containing an antigen binding site that specifically binds to an antigen (immunologically reacts with)). By "specifically binds to" or "immunologically reacts with" it is meant that the antibody reacts with one or more antigenic determinants of the desired antigen and does not react with other polypeptides. Each heavy chain is comprised of a heavy chain variable region (abbreviated herein as HCVR or VH) and at least one heavy chain constant region. Each light chain is comprised of a light chain variable region (abbreviated herein as LCVR or VL) and a light chain constant region. The variable regions of the heavy and light chains contain the binding domain that interacts with an antigen. The VHand VLregions can be further subdivided into regions of hypervariability, termed complementarity determining regions (CDRs), interspersed with regions that are more conserved, termed framework regions (FRs).

[0179] Antibodies can include, but are not limited to, polyclonal, monoclonal, chimeric, dAb (domain antibody), single chain, Fab, Fab' and F(ab')2 fragments, scFv, and Fab expression library.

[0180] The antibodies of the application can be monoclonal antibodies. Monoclonal antibodies (mAbs) of the application can be produced by various techniques, including conventional monoclonal antibody methodology e.g. those disclosed in “Monoclonal Antibodies: a manual of techniques” (Zola H, 1987, CRC Press) and “Monoclonal Hybridoma Antibodies: techniques and applications” (Hurrell JGR, 1982 CRC Press). Antibodies of the application can be multispecific, such as bispecific. Bispecific antibodies of the application bind to two different epitopes. These epitopes can be in the same protein (e.g. two epitopes in the spike protein of SARS-CoV-2) or different proteins (e.g. one epitope in the spike protein of SARS-CoV-2 and one epitope in another protein such as the coat protein).

[0181] In one embodiment, the bispecific antibodies of the application can bind to two separate epitopes on the spike protein of SARS-CoV-2. The bispecific antibodies can bind to the NTD of the spike protein and the RBD of the spike protein. The bispecific antibodies can bind to two different epitopes in the RBD of the spike protein.

[0182] One or more (e.g. two) antibodies of the application can be coupled to form a multispecific (e.g. bispecific) antibody. Methods of making multispecific (e.g. bispecific) antibodies are well known in the art.

[0183] The antibody can be selected from the group consisting of single chain antibodies, single chain variable fragments (scFv), variable fragments (Fv), fragment antigen binding regions (Fab), recombinant antibodies, monoclonal antibodies, fusion proteins comprising the antigen binding domain of a natural antibody or aptamer, single domain antibodies (sdAb) (also known as VHH antibodies), nanobodies (single domain antibodies of camelid origin), single domain antibody fragments of shark IgNAR origin (known as VNAR), diabodies, triabodies, Anticalins, aptamers (DNA or RNA) and active components or fragments thereof.

[0184] The constant region domain of the antibody molecule of the application (if present) can be selected with consideration of the intended function of the antibody molecule, particularly the effector function that can be desired. For example, the constant region domain can be a human IgA, IgD, IgE, IgG, or IgM domain. Typically, the constant region is of human origin. In particular, a human IgG (i.e., IgGl, IgG2, IgG3, or IgG4) constant region domain can be used. Typically, the constant region is a human IgGl constant region.

[0185] The light chain constant region can be lambda or kappa.

[0186] The antibody of the application can be monospecific or multispecific (e.g., bispecific). A multispecific antibody comprises at least two different variable domains, wherein each variable domain is capable of binding to a separate antigen or to a different epitope on the same antigen.

[0187] The antibody of the application can be a chimeric antibody, a CDR-grafted antibody, a nanobody, a human antibody, or a humanized antibody. Typically, the antibody is a human antibody. Fully human antibodies are those in which the variable and constant regions of the heavy and light chains are of human origin or are substantially identical to sequences of human origin but not necessarily from the same antibody.

[0188] The antibody of the application can be a full-length antibody.

[0189] The antibody of the application can be an antigen-binding fragment. An antigen-binding fragment of the application binds to the same epitope as the parent antibody from which the antigen-binding fragment is derived. An antigen-binding fragment of the application typically retains the portion of the parent antibody that interacts with the epitope. An antigen-binding fragment typically comprises a complementarity determining region (CDR) that interacts with the antigen, such as one, two, three, four, five, or six CDRs. In some embodiments, an antigen-binding fragment further comprises a structural scaffold that surrounds the CDRs of the parent antibody, such as the variable region domain of the heavy and / or light chain. Typically, an antigen-binding fragment retains the same or similar binding affinity for the antigen as the parent antibody.

[0190] An antigen-binding fragment does not necessarily have the same sequence as the parent antibody. In one embodiment, an antigen-binding fragment can have >70%, >80%, >90%, >95%, >96%, >97%, >98%, >99%, 100% sequence identity to the corresponding CDRs of the parent antibody. In one embodiment, an antigen-binding fragment can have >70%, >80%, >90%, >95%, >96%, >97%, >98%, >99%, 100% sequence identity to the corresponding variable region domain of the parent antibody. Typically, the non-identical amino acids of the variable region are not in the CDRs.

[0191] Antigen binding fragments of the antibodies of the application retain the ability to bind selectively to an antigen. Antigen binding fragments of the antibodies include single chain antibodies (i.e., full length heavy and light chains); Fab, modified Fab, Fab', modified Fab', F(ab')2, Fv, Fab-Fv, Fab-dsFv, single domain antibodies (e.g., VH or VL or VHH), scFv.

[0192] The antigen binding function of an antibody can be achieved by fragments of full length antibodies. Methods for producing and manufacturing these antibody fragments are well known in the art (see, e.g., Verma R et al., 1998, J. Immunol. Methods, 216, 165-181).

[0193] Methods for screening antibodies of the application having less than 100% amino acid sequence identity to one of the antibodies disclosed herein, having the desired specificity, affinity and functional activity include those described herein, e.g., enzyme-linked immunosorbent assay, biacore, focus reduction neutralization test (FRNT), and others known in the art.

[0194] With respect to function, the antibodies of the application can be capable of neutralizing at least one biological activity of SARS-CoV-2 (neutralizing antibodies), in particular neutralizing viral infectivity.

[0195] Neutralization can also be determined using IC50or IC90values. For example, the antibody can have an IC50value of < 0.1 pg / ml, < 0.05 pg / ml, < 0.01 pg / ml, < 0.005 pg / ml, or < 0.002 pg / ml. In some cases, the antibodies of the application can have an IC50value of between 0.0001 pg / ml and 0.1 pg / ml, sometimes between 0.0001 pg / ml and 0.05 pg / ml, or even between 0.0001 pg / ml and 0.001 pg / ml.

[0196] For example, IC50values for some of the antibodies of Tables 1-12 are provided in Tables 13-16.

[0197] The ability of an antibody to neutralize viral infectivity can be measured using an appropriate assay, in particular using a cell-based neutralization assay, as shown in the examples. For example, the neutralization ability can be measured in a focus reduction neutralization test (FRNT), in which the number of cells (e.g., human cells) infected by a virus in the presence of the antibody (e.g., infected for 2 hours at 37°C) is compared to a negative control in which no antibody was added.

[0198] The antibodies of the application can block the interaction between the spike protein of SARS-CoV-2 and the cell surface receptor angiotensin-converting enzyme 2 (ACE2) of a target cell, for example by direct blocking or by disrupting the prefusion conformation of the spike protein.

[0199] The blocking of the interaction between the spike protein and ACE2 can be complete or partial. For example, the antibodies of the application can reduce spike protein-ACE2 formation by >50%, >60%, >70%, >80%, >90%, >95%, >99% or 100%. The blocking of spike protein-ACE2 formation can be measured by any suitable means known in the art, for example, by ELISA.

[0200] Most antibodies that show neutralization also show blocking of the interaction between the spike protein and ACE2. Furthermore, many non-neutralizing antibodies are good ACE2 blockers.

[0201] In terms of binding kinetics, the antibodies of the application can have an affinity constant (KD) value for the spike protein of SARS-CoV-2 of <5 nM, <4 nM, <3 nM, <2 nM, <1 nM, <0.5 nM, <0.4 nM, <0.3 nM, <0.2 nM or <0.1 nM.

[0202] The KD value can be measured by any suitable means known in the art, for example, by ELISA or surface plasmon resonance (Biacore) at 25 °C.

[0203] Binding affinity (KD) can be quantified by determining the dissociation constant (Kd) and the association constant (Ka) of the antibody and its target. For example, the antibody can have an association constant (Ka) of >10000 M -1 s -1 , >50000 M -1 s -1 , >100000 M -1 s -1 , >200000 M -1 s -1 or >500000 M -1 s -1 and / or a dissociation constant (Kd) of <0.001 s -1 , <0.0005 s -1 , <0.004 s -1 , <0.003 s -1 , <0.002 s -1 or <0.0001 s -1 .

[0204] The antibodies of the application preferably are capable of providing in vivo protection in an animal infected with a coronavirus (e.g., SARS-CoV-2). For example, administration of an antibody of the application to an animal infected with a coronavirus (e.g., SARS-CoV-2) can result in a survival rate of >30%, >40%, >50%, >60%, >70%, >80%, >90%, >95%, or 100%. Survival rate can be determined using routine methods.

[0205] The antibodies of the application can have any combination of one or more of the above-mentioned properties.

[0206] The antibodies of the application can bind to the same epitope as any of the antibodies described herein (i.e., in particular, to the same epitope as an antibody having the heavy and light chain variable regions described above), or compete with any of the antibodies described herein for binding to the SARS-CoV-2 spike protein. Methods for identifying antibodies that bind to the same epitope or cross-compete with one another are used in the Examples and are discussed further below.

[0207] Fc region

[0208] The antibodies of the application can or can not comprise an Fc domain.

[0209] The antibodies of the application can be modified in the Fc region in order to improve their stability. Such modifications are known in the art. The modifications can improve the stability of the antibody during storage of the antibody. The in vivo half-life of the antibody can be improved by modifications to the Fc region. For example, a cysteine residue can be introduced into the Fc region, thereby allowing interchain disulfide bonds to form in this region. The homodimeric antibody thus produced can have improved internalization capability and / or increased complement-mediated cell killing and antibody-dependent cellular cytotoxicity (ADCC) (see Caron et al., J. Exp Med., 176:1191-1195 (1992) and Shopes, J. Immunol., 148:2918-2922 (1992)).

[0210] Alternatively, an antibody with dual Fc regions can be engineered, whereby complement lysis and ADCC capabilities can be enhanced (see Stevenson et al., Anti-Cancer Drug Design, 3:219-230 (1989)).

[0211] For example, the antibodies of the application can be modified to promote the interaction of the Fc domain with FcRn. The Fc domain can be modified to improve the stability of the antibody by affecting the Fc and FcRn interaction at low pH, such as in endosomes. The M252Y / S254T / T256E (YTE) mutation can be used to improve the half-life of IgG1 antibodies.

[0212] Antibodies can be modified to affect the interaction of the antibody with other receptors, such as FcyRI, FcyRIIA, FcyRIIB, FcyRIII, and FcaR. Such modifications can be used to affect the effector function of the antibody.

[0213] In one embodiment, the antibodies of the application comprise an altered Fc domain as described below. In another preferred embodiment, the antibodies of the application comprise an Fc domain, but the sequence of the Fc domain has been altered to modify one or more Fc effector functions.

[0214] In one embodiment, the antibodies of the application comprise a“silent” Fc region. For example, in one embodiment, the antibodies of the application do not exhibit effector function or functions associated with a normal Fc region. The Fc region of the antibodies of the application does not bind to one or more Fc receptors.

[0215] In one embodiment, the antibodies of the application do not comprise a CH2 domain. In one embodiment, the antibodies of the application do not comprise a CH3 domain. In one embodiment, the antibodies of the application comprise additional CH2 and / or CH3 domains.

[0216] In one embodiment, the antibodies of the application do not bind to Fc receptors. In one embodiment, the antibodies of the application do not bind to complement. In an alternative embodiment, the antibodies of the application do not bind to FcyR, but do bind to complement.

[0217] In one embodiment, the antibodies of the application can comprise modifications that alter the serum half-life of the antibody. Accordingly, in another embodiment, the antibodies of the application have Fc region modifications that alter the half-life of the antibody. Such modifications can be present in addition to those that alter Fc function. In a preferred embodiment, the antibodies of the application have modifications that alter the serum half-life of the antibody.

[0218] In one embodiment, the antibodies of the application can comprise human constant regions, for example, IgA, IgD, IgE, IgG, or IgM domains. In particular, when the antibody molecule is intended for therapeutic use where antibody effector functions are desired, human IgG constant region domains, especially IgGl and IgG3 isotypes, can be used. Alternatively, when the antibody molecule is for therapeutic purposes and antibody effector functions are not desired, IgG2 and IgG4 isotypes can be used.

[0219] In one embodiment, the antibody heavy chain comprises a CH1 domain, and the antibody light chain comprises a CL domain (kappa or lambda). In one embodiment, the antibody heavy chain comprises a CH1 domain, a CH2 domain, and a CH3 domain, and the antibody light chain comprises a CL domain (kappa or lambda).

[0220] The four human IgG isotypes bind the activating Fcy receptors (FcyRI, FcyRIIa, FcyRIIc, FcyRIIIa), the inhibitory FcyRIIb receptor and the first component of complement (Clq) with different affinities, resulting in very different effector functions (Brhns P. et al. 2009. Specificity and affinity of human Fcy receptors and their polymorphic variants for human IgG subclasses. Blood. 113(16):3716-25), see also Jeffrey B. Stavenhagen et al. Cancer Research September 15, 2007; 67(18): 8882-90). In one embodiment, the antibody of the application does not bind to an Fc receptor. In another embodiment of the application, the antibody does bind to one or more types of Fc receptor.

[0221] In one embodiment, the Fc region employed is mutated, in particular to the mutations described herein. In one embodiment, the Fc mutations are selected from the group comprising mutations that remove or enhance Fc region binding to Fc receptors, mutations that increase or remove effector functions, mutations that increase or decrease the half-life of the antibody, and combinations thereof. In one embodiment, the impact of a modification can be demonstrated by comparison with an equivalent antibody lacking the modification.

[0222] Some antibodies that selectively bind FcRn at pH 6.0 but not at pH 7.4 exhibit longer half-lives in various animal models. Several mutations located at the interface between the CH2 and CH3 domains (such as T250Q / M428L (Hinton PR. et al., 2004. Engineered human IgG antibodies with longer serum half-lives in primates. J Biol Chem. 279(8):6213-6) and M252Y / S254T / T256E + H433K / N434F (Vaccaro C. et al., 2005. Engineering the Fc region of immunoglobulin G to modulate in vivo antibody levels. Nat Biotechnol. 23(10): 1283-8)) have been shown to increase binding affinity for FcRn and half-life of IgGl in vivo. Thus, modifications that alter serum half-life can be present at M252 / S254 / T256 + H44 / N434, particularly M252Y / S254T / T256E + H433K / N434F can be present. In one embodiment, it is desirable to increase half-life. In another embodiment, it can actually be desirable to decrease the serum half-life of the antibody, thus modifications that decrease serum half-life can be present.

[0223] A number of mutations have been made in the CH2 domain of human IgGl and tested in vitro for their effect on ADCC and CDC (Idusogie EE. et al., 2001. Engineered antibodies with increased activity to recruit complement. J Immunol. 166(4):2571-5). Of note, an alanine substitution at position 333 was reported to increase ADCC and CDC. Thus, in one embodiment, a modification can be present at position 333, particularly one that alters the ability to recruit complement. Lazar et al. describe a triple mutant (S239D / I332E / A330L) that has a higher affinity for FcyRIIIa and a lower affinity for FcyRIIb, thereby enhancing ADCC (Lazar GA. et al., 2006). Thus, modifications can be present at S239 / I332 / A330, particularly those that alter affinity for Fc receptors, particularly S239D / I332E / A330L. Engineered antibody Fc variants with enhanced effector function. PNAS 103(11):4005-4010). The same mutations were used to generate an antibody with increased ADCC (Ryan MC. et al., 2007. Antibody targeting of B-cell maturation antigen on malignant plasma cells. Mol. Cancer Ther, 6:3009-3018). Richards et al. investigated a slightly different triple mutant (S239D / I332E / G236A) with improved FcyRIIIa affinity and FcyRIIa / FcyRIIb ratio that can mediate enhanced macrophage phagocytosis of target cells (Richards JO et al., 2008. Optimization of antibody binding to Fcgamma RIIa enhances macrophage phagocytosis of tumor cells. Mol Cancer Ther. 7(8):2517-27). In one embodiment, thus the S239D / I332E / G236A modification can be present.

[0224] In another embodiment, the antibodies of the application can have a modified hinge region and / or CH1 region. Alternatively, the isotype employed can be selected because it has a particular hinge region.

[0225] Major public V regions

[0226] A public V region (also described herein as a public V gene) is a V region of a germline heavy and light chain region that is found in a substantial portion of the antibody response to SARS-CoV-2 in a population. In the present application, the V region specifically reacts with the Beta SARS-CoV-2 variant. That is, many individuals utilize the same v region from their repertoire of germline v regions when mounting an immune response to SARS-CoV-2 variants.

[0227] As used herein, an antibody that is “derived from” a particular v region refers to an antibody that is produced by V(D)J recombination using that germline v region sequence. For example, a germline IGHV3-53 v region sequence can undergo somatic recombination and somatic mutation to produce an antibody that specifically binds to the spike protein of SARS-CoV-2. The nucleotide sequence encoding this antibody is unlikely to comprise the same sequence as the IGHV3-53 germline sequence, yet the antibody is still derived from this v region. Antibodies of the application typically comprise only non-silent mutations in the v region when compared to the germline sequence, such as no more than 15, 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 non-silent mutations. Antibodies of the application typically do not comprise 2-20 non-silent mutations in the v region when compared to the germline sequence, such as 5-15, 6-13, and 7-12 non-silent mutations. Germline v region sequences are well known in the art, and methods of identifying whether a certain region of an antibody is derived from a particular germline v region sequence are also well known in the art.

[0228] In one embodiment, the antibodies of the application are derived from a v region selected from IGHV3-53, IGHVl-58, IGHV3-66, IGHVl-69, IGHV3-30, IGHV3-33, IGHVl-18, IGHV13-9, or IGHV4-31. The inventors have found that the potent neutralizing antibodies identified herein comprise relatively few mutations in the CDRs of these v regions. Thus, in one embodiment, the antibodies of the application are encoded by a v region selected from IGHV3-53, IGHVl-58, IGHV3-66, IGHVl-69, IGHV3-30, IGHV3-33, IGHVl-18, IGHV13-9, or IGHV4-31, and have 2-20 non-silent nucleotide mutations or 5-15 non-silent mutations when compared to the naturally occurring germline sequence, such as 15 or fewer, 14 or fewer, 13 or fewer, 12 or fewer, 11 or fewer, 10 or fewer, 9 or fewer, 8 or fewer, 7 or fewer, 6 or fewer, 5 or fewer, 4 or fewer, 3 or fewer, or 2 non-silent mutations. A silent mutation, as defined herein, is a change in the nucleotide sequence that does not change the amino acid sequence encoded by that nucleotide sequence. Thus, a non-silent mutation is a mutation that results in a change in the amino acid sequence encoded by the nucleotide sequence.

[0229] The inventors have surprisingly found that the light chain variable regions of two antibodies having the same heavy chain v region can be swapped to produce a hybrid chain antibody comprising the heavy chain variable region of a first antibody and the light chain variable region of a second antibody. For example, both antibodies can comprise a heavy chain variable region derived from IGHV3-53. Preferably, both antibodies also comprise a light chain variable region derived from the same light chain v region, although this is not necessary, as the light chain of antibody 222, for example, can be paired with any heavy chain variable region derived from IGHV3-53 and produce a potent neutralizing antibody. As described above, both antibodies can comprise a heavy chain variable region derived from IGHV3-53 and / or IGHV3-66.

[0230] In one embodiment, the antibody of the application comprises CDRs derived from the heavy chain variable domain of an antibody selected from the group consisting of IGHV3-53, IGHV1-58, IGHV3-66, IGHV4-39, IGHV3-30, IGHV5-51, IGHV1-02, or IGHV3-33, such as antibody Omi03, Omi18, Omi29, Beta-27, antibody 150, antibody 158, antibody 175, antibody 222, and antibody 269 for IGHV3-53, antibody Omi16, Omi17, Omi20, Omi27, Omi36, antibody 40, and antibody 398 for IGHV3-66, antibody Omi12, Beta-47, Beta-25, antibody 55, antibody 165, antibody 253 for IGHV1-58, antibody Beta-49, Beta-50, Omi02, Omi24, Omi30, Omi31, Omi34, and Omi38 for IGHV1-69, antibody Beta-22, Beta-29, antibody 159, and Omi09 for IGHV3-30, antibody Beta-20, Beta-43, Omi32, and Omi33 for IGHV3-33, antibody 278, Beta-44, Omi26, and Omi41 for IGHV1-18, antibody 58, Omi25, Omi35, and Omi42 for IGHV3-9, or antibody Beta-56 and Omi23 for IGHV4-31. The SEQ ID NOs corresponding to the CDRs of each of these antibodies are shown in Tables 1, 2, and 3.

[0231] In one embodiment, the antibody of the application comprises a heavy chain variable domain derived from an antibody selected from the group consisting of IGHV3-53, IGHV1-58, IGHV3-66, IGHV4-39, IGHV3-30, IGHV5-51, IGHV1-02, or IGHV3-33, such as antibody Omi03, Omi18, Omi29, Beta-27, antibody 150, antibody 158, antibody 175, antibody 222, and antibody 269 for IGHV3-53, antibody Omi16, Omi17, Omi20, Omi27, Omi36, antibody 40, and antibody 398 for IGHV3-66, antibody Omi12, Beta-47, Beta-25, antibody 55, antibody 165, antibody 253 for IGHV1-58, antibody Beta-49, Beta-50, Omi02, Omi24, Omi30, Omi31, Omi34, and Omi38 for IGHV1-69, antibody Beta-22, Beta-29, antibody 159, and Omi09 for IGHV3-30, antibody Beta-20, Beta-43, Omi32, and Omi33 for IGHV3-33, antibody 278, Beta-44, Omi26, and Omi41 for IGHV1-18, antibody 58, Omi25, Omi35, and Omi42 for IGHV3-9, or antibody Beta-56 and Omi23 for IGHV4-31. The SEQ ID NOs corresponding to the CDRs of each of these antibodies are shown in Tables 1, 2, and 3.

[0232] In one embodiment, the application provides a method of generating an antibody that specifically binds to a spike protein of SARS-CoV-2 (e.g., a SARS-CoV-2 strain of the Alpha, Beta, Gamma, Delta, and / or Omicron lineage), the method comprising identifying two or more antibodies derived from the same light chain and / or heavy chain v-region, replacing the light chain of a first antibody with the light chain of a second antibody, thereby generating a hybrid chain antibody comprising the heavy chain of the first antibody and the light chain of the second antibody. In one embodiment, the method further comprises determining the affinity and / or neutralization of the hybrid chain antibody for SARS-CoV-2. The method can further comprise comparing the affinity of the hybrid chain antibody to the affinity of the first and / or second antibody. The method can further comprise selecting the hybrid chain antibody that has the same affinity or greater affinity than the first and / or second antibody. In some embodiments, the heavy chain v-region is IGHV 1-58, and / or the light chain v-region is IGLV Kappa 3-20.

[0233] In another embodiment, the present application provides an antibody that specifically binds to the Omicron variant of SARS-CoV-2, wherein the antibody has a v-region derived from IGHV1-69. It was surprisingly found that antibody responses to the Omicron variant of SARS-CoV-2 are biased towards antibodies having a heavy chain variable region derived from IGHV1-69. In one embodiment, wherein the antibody heavy chain is derived from IGHV1-69, the antibody of the present application comprises the CDRH1, CDRH2, and CDRH3 from Beta-49, Beta-50, Omi02, Omi24, Omi30, Omi31, Omi34, and Omi38.

[0234] Antibody conjugates

[0235] The present application also relates to immunoconjugates comprising an antibody conjugated to a cytotoxic agent such as a toxin (e.g., an enzymatically active toxin of bacterial, fungal, plant, or animal origin, or fragments thereof), or a radioisotope (i.e., a radioconjugate). Conjugates of an antibody and a cytotoxic agent can be made using a variety of bifunctional protein coupling agents

[0236] An antibody of the present application can be conjugated to a molecule that modulates or alters serum half-life. An antibody of the present application can be bound to albumin, for example, in order to modulate serum half-life. In one embodiment, an antibody of the present application will further comprise a binding region specific for albumin. In another embodiment, an antibody of the present application can comprise a peptide linker that is an albumin binding peptide. Examples of albumin binding peptides include those in WO 2015 / 197772 and WO 2007 / 106120, the entire contents of which are incorporated by reference.

[0237] Polynucleotides, vectors, and host cells

[0238] The present application also provides one or more isolated polynucleotides (e.g., DNA) encoding an antibody of the present application. In one embodiment, the polynucleotide sequences are collectively present on more than one polynucleotide, but together they are collectively capable of encoding an antibody of the present application. For example, the polynucleotide can encode the heavy chain and / or light chain variable region of an antibody of the present application. The polynucleotide can encode the complete heavy chain and / or light chain of an antibody of the present application. Typically, one polynucleotide will encode each of the heavy chain and light chain.

[0239] Polynucleotides encoding the antibodies of the application can be obtained by methods well known to those of skill in the art. For example, DNA sequences encoding part or all of the antibody heavy and light chains can be synthesized as desired from the corresponding amino acid sequences. General methods of constructing vectors, methods of transfection, and methods of culturing are well known to those of skill in the art. In this regard, reference is made to "Current Protocols in Molecular Biology", 1999, F. M. Ausubel (ed.), Wiley Interscience, New York and the Maniatis manual published by Cold Spring Harbor Press. The polynucleotides of the application can be provided in the form of expression cassettes comprising control sequences operably linked to the insert sequences, thereby allowing expression of the antibodies of the application in vivo. Accordingly, the application also provides one or more expression cassettes encoding one or more polynucleotides encoding an antibody of the application. These expression cassettes are in turn typically provided within a vector (e.g., a plasmid or a recombinant viral vector). Accordingly, in one embodiment, the application provides a vector encoding an antibody of the application. In another embodiment, the application provides vectors that collectively encode an antibody of the application. These vectors can be cloning vectors or expression vectors. A suitable vector can be any vector that is capable of carrying a sufficient amount of genetic information and allowing expression of a polypeptide of the application. The polynucleotides, expression cassettes, or vectors of the application are introduced into a host cell, for example by transfection. Accordingly, the application also provides a host cell comprising one or more polynucleotides, expression cassettes, or vectors of the application. The polynucleotides, expression cassettes, or vectors of the application can be introduced into the host cell transiently or permanently, thereby allowing expression of the antibody from the one or more polynucleotides, expression cassettes, or vectors. Such host cells include transient or, preferably, stable higher eukaryotic cell lines (such as mammalian cells or insect cells), lower eukaryotic cells (such as yeast), or prokaryotic cells (such as bacterial cells). Particular examples of cells include mammalian HEK293 (such as HEK293F, HEK293T, HEK293S, or HEK Expi293F), CHO, HeLa, NS0, and COS cells or any other cell line used herein (such as the cell lines used in the examples). Preferably, the cell line chosen will be one that is not only stable but also allows for mature glycosylation.

[0240] The application also provides a process for producing an antibody of the application, the process comprising culturing a host cell containing one or more vectors of the application under conditions suitable for expression of the antibody from the one or more polynucleotides of the application, and isolating the antibody from the culture.

[0241] Antibody combinations

[0242] The inventors have found that certain antibodies of Table 3 are particularly effective when used in combination, and that certain combinations of antibodies of Table 3, Table 2 and Table 1 maximise therapeutic efficacy and / or increase diagnostic capability (e.g. to minimise loss of activity due to SARS-CoV-2 variants). Useful combinations include antibodies that do not cross-compete and / or bind to non-overlapping epitopes.

[0243] The present application therefore provides an antibody combination of the present application, wherein each antibody is capable of binding to the spike protein of the coronavirus SARS-CoV-2, wherein at least one antibody comprises at least three CDRs of any one of the 28 antibodies of Table 3.

[0244] The antibody combination of the present application can be used as a therapeutic cocktail. The present application therefore also provides a pharmaceutical composition comprising an antibody combination of the present application, as explained further below.

[0245] The antibody combination of the present application can be used for diagnosis. The present application therefore also provides a diagnostic kit comprising an antibody combination of the present application. Methods of diagnosing a disease or complication associated with coronavirus infection in a subject are also provided herein, as explained further below. A fully cross-neutralising antibody (e.g. Omi03) can be used as a reference to confirm the presence and / or amount of any variant of concern (VoC) SARS-CoV-2 in a sample. An antibody that binds to a limited number of VoCs can be used to confirm the presence and / or amount of that VoC in a sample. For example, if Omi03 exhibits binding to a sample, but Omi24 does not exhibit binding to the sample of SARS-CoV-2, then the spike protein is likely to be that of the Delta VoC. This can be determined by any method known to the skilled person, such as via an immunoassay, e.g. an ELISA or an immunochromatographic assay. Reduced binding can be determined by comparison to a reference and / or normalisation to a reference, and / or by comparison to positive / negative control samples or data.

[0246] Pharmaceutical compositions

[0247] The present application provides a pharmaceutical composition comprising an antibody of the present application. The composition can comprise a combination (such as two, three or four) of antibodies of the present application. The pharmaceutical composition can further comprise a pharmaceutically acceptable carrier.

[0248] The composition of the present application can include one or more pharmaceutically acceptable salts. By “pharmaceutically acceptable salt” is meant a salt that retains the desired biological activity of the parent compound and does not impart any undesired toxicological effects. Examples of such salts include acid addition salts and base addition salts.

[0249] Suitable pharmaceutically acceptable carriers include aqueous carriers or diluents.

[0250] Examples of suitable aqueous vehicles include water, buffered water, and saline.

[0251] Other suitable pharmaceutically acceptable vehicles include ethanol, polyols (such as glycerol, propylene glycol, polyethylene glycol, and the like), and suitable mixtures thereof, vegetable oils, such as olive oil, and injectable organic esters, such as ethyl oleate. In many cases, it will be desirable to include isotonic agents (e.g., sugars), polyalcohols such as mannitol, sorbitol, or sodium chloride in the composition.

[0252] Therapeutic compositions typically must be sterile and stable under the conditions of manufacture and storage. The composition can be formulated as a solution, microemulsion, liposome, or other ordered structure suitable to high drug concentration.

[0253] The pharmaceutical compositions of the present application can comprise an additional therapeutic agent, for example an antiviral agent. The antiviral agent can bind to the coronavirus and inhibit viral activity. Alternatively, the antiviral agent can not bind directly to the coronavirus, but still affect viral activity / infectivity. The antiviral agent can be a further anti-coronavirus antibody, which binds somewhere on SARS-CoV-2 other than the spike protein. Examples of antiviral agents which can be used in the present application include Remdesivir, Lopinavir, ritonavir, APN01 and Favilavir.

[0254] The additional therapeutic agent can be an anti-inflammatory agent, such as a corticosteroid (e.g. dexamethasone) or a non-steroidal anti-inflammatory drug (e.g. Tocilizumab).

[0255] The additional therapeutic agent can be an anti-coronavirus vaccine. The pharmaceutical composition can be administered subcutaneously, intravenously, intradermally, intramuscularly, intranasally or orally. Also within the scope of the present application is a kit comprising an antibody or other composition of the present application and instructions for use. The kit can also contain one or more additional agents, such as an additional therapeutic or prophylactic agent as discussed herein.

[0256] Methods and uses of the invention

[0257] The present application also relates to the use of an antibody, antibody combination and pharmaceutical composition as described herein, for example in a method of treatment of the human or animal body by therapy or in a diagnostic method. The method of treatment can be therapeutic or prophylactic.

[0258] For example, the present application relates to a method of treating a coronavirus (e.g., SARS-CoV-2) infection, a disease or complication associated therewith (e.g., COVID-19). The method can comprise administering a therapeutically effective amount of an antibody, antibody combination or pharmaceutical combination of the present application. The method can further comprise identifying the presence of a coronavirus or fragment thereof (e.g., SARS-CoV-2) in a sample from a subject. The present application also relates to an antibody, antibody combination or pharmaceutical composition according to the present application for use in a method of treating a coronavirus (e.g., SARS-CoV-2) infection, a disease or complication associated therewith (e.g., COVID-19).

[0259] The present application also relates to a method of formulating a composition for treating a coronavirus (e.g., SARS-CoV-2) infection, a disease or complication associated therewith (e.g., COVID-19), wherein the method comprises mixing an antibody, antibody combination or pharmaceutical composition according to the present application with an acceptable carrier to make the composition.

[0260] The present application also relates to the use of an antibody, antibody combination or pharmaceutical composition according to the present application for treating a coronavirus (e.g., SARS-CoV-2) infection or a disease or complication associated therewith (e.g., COVID-19).

[0261] The present application also relates to the use of an antibody, antibody combination or pharmaceutical composition according to the present application for the manufacture of a medicament for treating or preventing a coronavirus (e.g., SARS-CoV-2) infection or a disease or complication associated therewith (e.g., COVID-19).

[0262] The present application also relates to the prevention, treatment or diagnosis of a coronavirus infection caused by any SARS-CoV-2 strain. The coronavirus infection can be caused by any SARS-CoV-2 strain.

[0263] SARS-CoV-2 strains can be the earliest identified ancestral strain ((hCoV-19 / Wuhan / WIV04 / 2019 (WIV04); GISAID Accession Number: EPI_ISL_402124) and variants thereof. For example, the SARS-CoV-2 strain can be a member of lineage A, A.1, A.2, A.3, A.5, B, B.1, B.1.1, B.2, B.3, B.4, B.1.1.7 (alpha), B.1.351 (beta), P.1 (gamma), delta, kappa, and / or lambda. The SARS-CoV-2 strain can be a member of lineage A.23.1, B.1.1.7 (alpha), B.1.351 (beta), B.1.258, B.1.526.2, B.1.616, B.1.617.1 (kappa), B.1.617.2 (delta), C36.3, C.37 (lambda), P.1 (gamma), B.1.1.529 (omicron), Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and / or Omicron BA.3.

[0264] The SARS-CoV-2 strain can comprise one or more mutations, for example in the spike protein, relative to hCoV-19 / Wuhan / WIV04 / 2019 (WIV04) (GISAID Accession Number: EPI_ISL_402124). In other words, the SARS-CoV-2 strain can be a modified hCoV-19 / Wuhan / WIV04 / 2019 (WIV04) strain comprising one or more modifications, for example in the spike protein.

[0265] The mutation can be a mutation (e.g., substitution) observed in the Omicron strain of SARS-CoV-2. Antibodies Omi02, Omi03, Omi12, Omi18, Omi28, Omi39, and Omi42 are particularly effective at neutralizing the Omicron SARS-Cov-2 strain. Accordingly, the present application can involve the use of these antibodies for the treatment, prevention, or diagnosis of a coronavirus infection caused by a SARS-Cov-2 strain.

[0266] The methods and uses of the present application can comprise inhibiting a disease state (such as COVID-19), for example, arresting its development; and / or relieving a disease state (such as COVID-19), for example, causing the regression of the disease state until a desired endpoint is reached.

[0267] The methods and uses of the application can comprise ameliorating or reducing the severity of symptoms of a disease state (such as COVID-19), improving or reducing the duration or frequency of (e.g. reducing pain or discomfort), and such amelioration can or can not be directly related to the disease. Symptoms or complications can be fever, headache, fatigue, loss of appetite, myalgia, diarrhea, vomiting, abdominal pain, dehydration, respiratory infection, cytokine storm, acute respiratory distress syndrome (ARDS) sepsis and / or organ failure (e.g. heart, kidney, liver, GI, lungs).

[0268] The methods and uses of the application can result in a reduction in viral load of a coronavirus (e.g. SARS-CoV-2), for example a reduction of >10%, >20%, >30%, >40%, >50%, >60%, >70%, >80%, >90% or 100% compared to prior to treatment. Methods of determining viral load are well known in the art, for example infection assays.

[0269] The methods and uses of the application can comprise preventing the occurrence of a coronavirus infection in a subject (e.g. a human), particularly when the subject is predisposed to complications associated with coronavirus infection.

[0270] The application also relates to identifying a subject infected with a coronavirus (such as SARS-CoV-2). For example, the methods and uses of the application can involve identifying the presence of a coronavirus (e.g. SARS-CoV-2) or a protein or fragment thereof in a sample. The detection can be performed in vitro or in vivo. In certain embodiments, the application involves population screening.

[0271] The application relates to identifying any SARS-CoV-2 strain, as described herein. The application can also relate to a method of identifying an escape mutant of SARS-CoV-2, the method comprising contacting a sample with the antibody combination of the application and identifying whether each antibody binds to the virus. The term “escape mutant” refers to a variant of SARS-CoV-2 comprising non-silent mutations that can affect the efficacy of existing treatments for SARS-CoV-2 infection. Typically, the non-silent mutations are in an epitope recognized by an existing art antibody and / or an antibody described herein that specifically binds to an epitope of SARS-CoV-2, for example in the spike protein of SARS-CoV-2. If the antibody does not bind to the target, it can indicate that the target comprises mutations that can alter the efficacy of existing treatments for SARS-CoV-2.

[0272] The methods and uses of the application can comprise contacting a sample with an antibody or antibody combination of the application and detecting the presence or absence of an antibody-antigen complex, wherein the presence of an antibody-antigen complex is indicative of a subject being infected with SARS-CoV-2.

[0273] Methods of determining the presence of antibody-antigen complexes are known in the art. For example, in vitro detection techniques include enzyme-linked immunosorbent assay (ELISA), Western blot, immunoprecipitation, and immunofluorescence. In vivo techniques include introducing a labeled anti-analyte protein antibody into a subject. For example, the antibody can be labeled with a radioactive marker, the presence and location of which in the subject can be detected by standard imaging techniques. Detection techniques can provide a qualitative or quantitative readout, depending on the assay employed.

[0274] In general, the present application relates to methods and uses for human subjects in need thereof. However, non-human animals are also envisaged, such as rats, rabbits, sheep, pigs, cows, cats or dogs. The subject can be at risk of exposure to a coronavirus infection, such as a healthcare worker or someone who has been in contact with an infected individual. The subject can have visited or plan to visit a country where a coronavirus outbreak is known or suspected. The subject can also be at greater risk, such as an immunocompromised individual, for example, an individual receiving immunosuppressive therapy or an individual with human immunodeficiency syndrome (HIV) or acquired immune deficiency syndrome (AIDS). The subject can be asymptomatic or pre-symptomatic.

[0275] The subject can be in the early, middle or late stages of the disease.

[0276] The subject can be in a hospital or community at the time of first onset and / or later in the hospital.

[0277] The subject can be male or female.

[0278] In certain embodiments, the subject is typically male. The subject can not be infected with a coronavirus, such as SARS-CoV-2. The subject can be predisposed to more severe symptoms or complications associated with coronavirus infection. The method or use of the present application can comprise the step of identifying whether the patient is at risk of developing more severe symptoms or complications associated with coronavirus.

[0279] In embodiments where the present application relates to prevention or treatment, the subject can or can not have been diagnosed with infection with a coronavirus, such as SARS-CoV-2.

[0280] The present application relates to analyzing a sample from a subject. The sample can be tissue, cells and biological fluids isolated from the subject, as well as tissue, cells and fluids present in the subject’s body. The sample can be blood and a part or component of blood, including serum, plasma or lymph. Typically, the sample is from a throat swab, nasal swab or saliva.

[0281] The antibody-antigen complex detection assay can be performed in situ, in which case the sample is a tissue section (fixed and / or frozen) of tissue obtained from a biopsy or resection of the subject.

[0282] In embodiments of the application, where antibody pharmaceutical compositions and combinations are administered, they can be administered subcutaneously, intravenously, intradermally, orally, intranasally, intramuscularly, or intracranially. Typically, antibody pharmaceutical compositions and combinations are administered intravenously or subcutaneously.

[0283] The dosage of the antibody can vary depending on the age and size of the subject and the disease, condition, and route of administration. The antibody can be administered at a dosage of about 0.1 mg / kg body weight to about 100 mg / kg body weight, such as at a dosage of about 5 mg / kg to about 10 mg / kg. The antibody can also be administered at a dosage of about 50 mg / kg, 10 mg / kg, or about 5 mg / kg body weight.

[0284] The combinations of the application can be administered, for example, at a dosage of about 5 mg / kg to about 10 mg / kg of each antibody or at a dosage of about 10 mg / kg or about 5 mg / kg of each antibody. Alternatively, the combination can be administered at a total dosage of about 5 mg / kg (e.g., in a three-antibody combination, a dosage of 1.67 mg / kg of each antibody).

[0285] The antibodies or antibody combinations of the application can be administered in a multiple dosage regimen. For example, a second or multiple subsequent doses can be administered following an initial dose. The second and subsequent doses can be spaced at appropriate times.

[0286] As discussed above, the antibodies of the application are typically used in a single pharmaceutical composition / combination (co-formulation). However, the application generally also includes the use of the antibodies of the application in separate formulations / compositions. The application also includes the use of the antibodies as described above in combination with additional therapeutic agents.

[0287] The administration of combinations of two or more agents and / or antibodies can be achieved in a number of different ways. In one embodiment, all components can be administered together in a single composition. In another embodiment, each component can be administered separately as part of a combination therapy.

[0288] For example, the antibodies of the application can be administered prior to, after, or simultaneously with another antibody or binding fragment thereof of the application. Particularly useful combinations are described above, for example.

[0289] For example, the antibodies of the application can be administered prior to, after, or simultaneously with an antiviral agent or an anti-inflammatory agent.

[0290] In embodiments where the application relates to detecting the presence of a coronavirus (e.g. SARS-CoV-2) or a protein or fragment thereof in a sample, the antibody contains a detectable label. Methods of attaching labels to antibodies are known in the art, for example direct labelling of the antibody by coupling (i.e. physically linking) a detectable substance to the antibody. Alternatively, the antibody can be indirectly labelled, for example by reaction with an antibody which itself is directly labelled. Examples of indirect labelling include the use of a secondary antibody which is directly labelled with a fluorescent marker to detect the primary antibody, and the use of biotinylated DNA probes which can be end-labelled with biotin so that they can be detected with fluorescently labelled streptavidin.

[0291] The detection can also comprise (i) an agent known to be useful in detecting a coronavirus (e.g. SARS-CoV-2) or a protein or fragment thereof, for example an antibody to a further epitope of the spike protein or a further protein of the coronavirus, such as an anti-nucleocapsid antibody; and / or (ii) an agent known not to detect the presence of a coronavirus (e.g. SARS-CoV-2) or a fragment thereof, i.e. to provide a negative control.

[0292] In certain embodiments, the antibody is modified to have increased stability. Suitable modifications are explained above.

[0293] The application also includes a kit for detecting the presence of a coronavirus (e.g. SARS-CoV-2) in a sample. For example, the kit can comprise a labelled antibody or combination of labelled antibodies of the application; means for determining the amount of coronavirus (e.g. SARS-CoV-2) in the sample; and means for comparing the amount of coronavirus (e.g. SARS-CoV-2) in the sample to a standard. The labelled antibody or combination of labelled antibodies can be packaged in a suitable container. The kit can also include instructions for using the kit to detect a coronavirus (e.g. SARS-CoV-2) in a sample. The kit can also include other agents known to be useful in detecting the presence of a coronavirus, as discussed above.

[0294] For example, the antibody or combination of antibodies of the application is used in a lateral flow test. Typically, a lateral flow test kit is a hand-held device with an absorbent pad that is based on a series of capillary beds, such as a porous paper strip, a microstructured polymer or a sintered polymer. The test runs a liquid sample along the surface of the pad with a reaction molecule that shows a visual positive or negative result. The test can also include the use of other agents known to be useful in detecting the presence of a coronavirus (e.g. SARS-CoV-2) or a fragment thereof, as discussed above, such as an anti-nucleocapsid antibody.

[0295] Other

[0296] It is understood that different applications of the disclosed antibody combinations or pharmaceutical compositions of the present application can be adjusted according to particular needs in the art. It is also understood that the terminology used herein is for the purpose of describing particular embodiments of the application only and is not intended to be limiting. Furthermore, as used in the specification and the appended claims, the singular forms "a," "an" and "the" include plural references unless the context clearly dictates otherwise. Thus, for example, reference to "an antibody" includes two or more antibodies.

[0297] Further, when reference is made herein to "≥ x", this means equal to or greater than x. When reference is made herein to "≤ x", this means less than or equal to x.

[0298] For purposes of the present application, to determine the percent identity of two sequences, such as two polynucleotide or two polypeptide sequences, the sequences are aligned for optimal comparison purposes (e.g., gaps can be introduced in the sequence of a first sequence for optimal alignment with a second sequence). The nucleotides or amino acid residues at each position are then compared. When a position in the first sequence is occupied by the same nucleotide or amino acid as the corresponding position in the second sequence, then the nucleotides or amino acid residues at that position are identical. The percent identity between two sequences is a function of the number of identical positions shared by the sequences (i.e., % identity = # of identical positions / total # of positions in reference sequence x 100). Generally, the sequence comparison is performed over the length of the reference sequence. For example, if a user desires to determine whether a given ("test") sequence is % identical to SEQ ID NO: 395, then SEQ ID NO: 3 would be the reference sequence. To assess whether a sequence is at least 95% identical to SEQ ID NO: 3 (an example of a reference sequence), the skilled artisan would align the length of SEQ ID NO: 3 and determine how many positions in the test sequence are identical to the positions of SEQ ID NO: 3. If at least 95% of the positions are identical, then the test sequence is at least 95% identical to SEQ ID NO: 3. If the sequence is shorter than SEQ ID NO: 3, then gap or missing positions should be considered non-identical positions. The skilled artisan is aware of different computer programs that can be used to determine homology or identity between two sequences. For example, the comparison of sequences and the determination of percent identity between two sequences can be accomplished using a mathematical algorithm. In one embodiment, the percent identity between two amino acid or nucleic acid sequences is determined using the Needleman and Wunsch (1970) algorithm which has been incorporated into the GAP program in the Accelrys GCG software package (available at http: / / www.accelrys.com / products / gcg / available) using either a Blosum 62 matrix or a PAM250 matrix, and a gap weight of 16, 14, 12, 10, 8, 6, or 4 and the gap length weight of 1, 2, 3, 4, 5, or 6.

[0299] According to the IMGT numbering system (http: / / www.imgt.org; Lefranc MP, 1997, J, Immunol. Today 18, 509), the CDRs of the heavy chain (CDRH) and light chain variable domains (CDRL) are located at residues 27-38 (CDR1), residues 56-65 (CDR2) and residues 105-117 (CDR3) of each chain. This numbering system is used in this specification, unless otherwise stated.

[0300] All publications, patents and patent applications cited herein, whether supra or infra, are hereby incorporated by reference in their entirety.

[0301] The following examples illustrate the application.

[0302] Examples

[0303] Example 1. Generation of specific antibodies against early pandemic SARS-CoV-2 and Beta SARS-CoV-2 strains

[0304] The antibodies in Table 1 relate to a group of mAbs raised against early pandemic strains of SARS-CoV-2. The antibodies in Table 2 relate to a group of mAbs raised against the Beta strain of SARS-CoV-2.

[0305] Further details of these antibodies can be found in International Application Nos. PCT / GB2022 / 050306 and PCT / GB2022 / 050307. Further information on the production and properties of these antibodies can be found in the following articles: Dejnirattisai, Wanwisa et al. “The antigenic anatomy of SARS-CoV-2 receptor binding domain.” Cell 184.8 (2021): 2183-2200. Supasa, Piyada et al. “Reduced neutralization of SARS-CoV-2 B.1.1.7 variant by convalescent and vaccine sera.” Cell 184.8 (2021): 2201-2211.

[0306] Liu, Chang et al. “The antibody response to SARS-CoV-2 Beta underscores the antigenic distance to other variants.” Cell host & microbe (2021).

[0307] Zhou, Daming et al. “Evidence of escape of SARS-CoV-2 variant B.1.351 from natural and vaccine-induced sera.” Cell 184.9 (2021): 2348-2361.

[0308] Dejnirattisai, Wanwisa et al. “Antibody evasion by the P.1 strain of SARS-CoV-2.” Cell 184.11 (2021): 2939-2954.

[0309] Liu, Chang et al. “Reduced neutralization of SARS-CoV-2 B.1.617 by vaccine and convalescent serum.” Cell 184.16 (2021): 4220-4236.

[0310] Dejnirattisai, Wanwisa et al. “SARS-CoV-2 Omicron-B.1.1.529 leads to widespread escape from neutralizing antibody responses.” Cell (2022).

[0311] Example 2. Generation of specific antibodies against Omicron strains of SARS-CoV-2

[0312] Omicron BA.2 lineage

[0313] Omicron BA.2 was first reported in South Africa on 17 November 2021, similar to the time of the BA.1 report. In many countries such as Denmark, India, and the UK, BA.2 has been increasing relative to BA.1 and now accounts for the majority of Omicron infections in Denmark, and there is growing evidence that BA.2 is more transmissible than BA.1, but no evidence of increased disease severity.

[0314] BA.2 is related to BA.1, sharing 21 amino acid substitutions throughout S, but there are also many differences. Compared to BA.2, BA.1 has an additional 6 amino acid deletions, 3 insertions, and 9 substitutions, and compared to BA.1, BA.2 has an additional 3 deletions and 7 substitutions. In the RBD, BA.1 contains the unique mutations S371L, G446S, and G496S, and in some isolates R346K (BA.1.1), while BA.2 carries S371F, T376A, D405N, and R408S. All of these residues have the potential to have different effects on antibody binding and can modulate neutralization, particularly BA.1 G446S, G496S, BA.2 D405N, R408S, which are at the edge of the ACE2 binding footprint, for BA.1.1, R346K changes are close to the N343 glycan and can modulate binding of potent antibodies to this region. BA.3 contains no unique mutations relative to BA.1 and BA.2 and appears to be a fusion of the two, being BA.1-like at the N-terminus and BA.2-like at the C-terminus by the mutation G496S.

[0315] Omicron lineages BA.4 and BA.5

[0316] In early April 2022, two new Omicron lineages were reported in Gauteng Province, South Africa, named BA.4 and BA.5. The S sequences of BA.4 and BA.5 are identical and are closely related to BA.2. The sequence diversity of Omicron S is shown in Figure 9 Compared to BA.2, BA.4 lacks residues 69 and 70 and contains 2 additional substitutions in the RBD: L452R and F486V. Finally, BA.4 lacks the change Q493R seen in BA.1 and BA.2, reverting to Q493 as in the Victoria / Wuhan strain.

[0317] Two additional mutations in the RBD are of most interest in antibody escape: L452R is a chemically radical change and one of a pair of changes in the Delta RBD (the other being T478K, already found in the Omicron lineage). The mutation F486L, found in sequences of SARS-CoV-2 isolated from mink early in the pandemic, is also an escape mutation site for several mAbs (Gobeil et al., 2021, “Effect of natural mutations of SARS-CoV-2 on spike structure, conformation, and antigenicity”. Science 373, 6555). The change F486V in BA.4 / 5 also reduces most of the hydrophobic side chains, similar to F486L, but more significantly. Residues 452 and 486 are both located near the edge of the ACE2 interacting surface (…). Figure 9 B), and these two residues, together with the reversion mutation to the ancestral sequence Q493 located within the ACE2 footprint, may regulate ACE2 affinity, as well as the neutralizing capacity of vaccines or naturally obtained serum. The L452R and F486V mutations may lead to more antibody escape, while the reversion mutation at 493 may reduce escape from early viral responses.

[0318] Omicron lineage BA.2.75

[0319] In early May 2022, a new Omicron BA.2 sublineage, named BA.2.75, was reported in India. It has since spread to several countries, including the UK, USA, Australia, Germany, and Canada. Compared to BA.2, BA.2.75 contains multiple mutations in the S protein, including four substitutions in the NTD (W152R, F157L, I210V, and G257S) and four substitutions in the RBD: D339H, G446S, N460K, and R493Q. Figure 16). The RBD mutations affect the main epitopes of neutralizing antibodies and have the potential to modulate ACE2 binding. D339H represents a further evolution of the G339D mutation found in all previous Omicron variants, which has been found to impair binding of certain ‘right-wing’ antibodies belonging to the IGHV1-69 family (e.g. Beta-49 and -50); it also belongs to the binding footprint of certain class 3 antibodies such as S309 / sotrovimab (Dejnirattisai et al., 2022; “SARS-CoV-2 Omicron-B.1.1.529 leads to widespread escape from neutralizing antibody responses.” Cell 185, 467-484 e415). G446S was found in BA.1, BA.1.1 and BA.3, but not in BA.2 and other BA.2 subvariants, and is also able to impair binding of certain class 3 antibodies that bind to the right shoulder such as REGN10987 / imdevimab (Dejnirattisai et al., 2022). The R493Q back mutation was also found in BA.4 / 5, which makes the virus more susceptible to neutralization by many class 1 and 2 antibodies that bind to the neck / left shoulder. This back mutation also increases the affinity for ACE2 (see below).

[0320] N460K is a new mutation not seen in previous VoCs or Omicron sublineages, but it was found after in vitro (yeast display) evolution of RBD-62, which has ultra-high ACE2 affinity (KD = 16-18 pM) (Dejnirattisai et al., 2022; Zahradnik et al., 2021 “SARS-CoV-2 variant prediction and antiviral drug design are enabled by RBD in vitro evolution.” Nat Microbiol 6, 1188-1198). Indeed, N460K causes a large increase in affinity for ACE2, second only to the effect of N501Y (Zahradnik et al., 2021). Furthermore, computational analysis predicts that N460K can affect the binding of certain antibodies belonging to the IGHV3-53 family (e.g. Omi-3) that have been shown to be able to potently neutralize all VoCs (Nutalai et al., 2022).

[0321] Using a neutralization assay, it was shown that Delta infection alone does not provide protection against BA.2.75 (no neutralization). Mutations in BA.2.75 compared to BA.2 lead to a decrease in neutralization titers of vaccine sera. Individual BA.2.75 mutations can lead to a greater decrease in neutralization titers compared to the complete BA.2.75 S sequence, but these mutations are balanced by the R393Q back mutation that has been selected to increase affinity to ACE2 and increase the transmissibility of BA.2.75. Further evolution of the Omicron lineage seems inevitable and there can be many possible trade-offs between antibody escape and ACE2 affinity that can and will be made, leading to successive waves of infection.

[0322] Emerging BA.2, BA.4 and BA.5 sublineages

[0323] Many lineages within the BA.2 and BA.5 clades are growing rapidly. Most notable is the high degree of convergent evolution, particularly at antigenic RBD positions such as 346, 444, 452, 460, 486, 490, 493, and 494. These lineages include examples from the BA.4 / 5 clade (containing L452R, F486V, and the back mutation R493Q), such as BA.4.6 and BF.7 (R346T), BA.4.7 (R346S), BQ.1 (K444T, N460K), and B0.1.1 (R346T, K444T, N460K); from the BA.2.75 clade (containing G339H, G446S, N460K, and the back mutation R493Q), BA.2.75.2 (R346T and F486S and BA.2.75 mutations), BN.1 (aka BA.2.75.5.1, with R346T, K356T, F490S, and BA.2.75 mutations), BM.1.1.1 (aka BA.2.75.3.1.1.1, with R346T, F486S, F490S, and BA.2.75 mutations). There are several other examples of second-generation BA.2 variants, such as BJ.1 (aka BA.2.10.1.1; G339H, R346T, L368I, V445P, G446S, V483A, and F490V), BA.2.10.4 (G446S, F486P, S494P, and R493Q back mutation), BS.1 (aka BA.2.3.2.1; R346T, L452R, N460K, G476S, and Q493R back mutation), BA.2.3.20 (K444R, N450D, L452M, N460K, E484R, and Q493R back mutation), and finally the BJ.1 x BM.1.1.1 (aka BA.2.75.3.1.1.1) recombinant XBB (which contains R346T, L368I, V445P, G446S, N460K, F486S, F490S, and Q493R back mutation relative to BA.2).

[0324] Outside of RBD, the degree of convergent evolution is less, but still present. Many second-generation BA.2 variant lineages contain deletions or mutations in the NTD, often similar to those seen in VoCs, such as Δ~144 in BJ.1, BS.1, and BA.2.10.4 (previously seen in Alpha and BA.1) and NSP12 G671S in BJ.1, XBB, and BA.2.10.4 (previously seen in Delta).

[0325] Potent neutralizing antibodies isolated following Omicron infection

[0326] Five volunteers who had recovered from a sequence-confirmed Omicron infection were recruited and sampled 10-14 days after symptom onset; all had received 2 doses of the Pfizer BioNtech vaccine prior to infection with Omicron. First, neutralisation assays were performed against Omicron BA.1 and Victoria, an early pandemic SARS-CoV-2 isolate containing only a single amino acid substitution in the S NTD (S247R) compared to the sequence of the ancestral strain used in all current vaccines. In all cases, the focus-reducing neutralisation 50% titre (FRNT50) for Omicron was higher than 100, but the titres were much lower than for Victoria at this early time point Figure 1 A).

[0327] B cells from the five donors were stained with full-length BA.1 trimer and single cells sorted by FACS Figure 1 B). After a degenerate RT-PCR reaction, heavy and light chain sequences were assembled into expression vectors using Gibson reaction and the products transfected into 293T cells. Culture supernatants were screened for reactivity against full-length BA.1 or wild-type S (WT Wuhan) as well as BA.1 RBD and NTD. A total of 1,122 single cells were sorted and 545 mAbs were recovered.

[0328] By ELISA, all mAbs were found to cross-react between WT and BA.1 S, indicating that they were likely produced by vaccine-induced memory B cells. In contrast to a previous set of monoclonal antibodies produced from primary cases of early pandemic infection (Dejnirattisai, Wanwisa et al. “The antigenic anatomy of SARS-CoV-2 receptor binding domain.” Cell 184.8 (2021): 2183-2200), a higher proportion of omicron-specific mAbs were found to react with the RBD (56%) compared to early pandemic mAbs (21%, P<0.0001) Figure 1 C). In addition, 129 of the 545 isolated mAbs bound to the BA.1 NTD.

[0329] Isolation of potent Omicron mAbs

[0330] All ELISA positive mAbs were subjected to neutralization assay and those showing the highest activity were selected for further study. The top 28 mAbs were selected for full characterization, all of which showed BA.1 FRNT50 titers < 100 ng / ml. Of the 28, 27 were found to bind RBD by ELISA (one, Omi-41, binds NTD), none cross-reacted with SARS-CoV-1 S protein.

[0331] Inspection of gene usage Figure 1 D, Table 17) showed that 9 of the 28 mAbs belonged to the VH3-53 and related VH3-66 gene family. VH3-53 and VH3-66 have been repeatedly isolated in SARS-CoV-2 infection, they form a public antibody response and bind to a site on the RBD neck and act to block ACE2 binding. It was previously observed that many VH3-53 and VH3-66 mAbs lose activity against the VoC containing the N501Y mutation, although some VH3-53 antibodies (mAbs 222 and Beta-27) were completely resistant to the N501Y change found in Alpha, Beta and Gamma, but activity against Omicron BA.1 or BA.2 was knocked down.

[0332] Approximately half of the gene families observed in the potent early pandemic antibodies (Table 1) were also present in the Omicron group Figure 1 C), perhaps the most striking difference is that VH1-69 was not a feature of the early antibodies, but was found in 6 of the 28 (2, 24, 30, 31, 34 and 38) in the potent Omicron group, and we also found it in 2 Beta antibodies (Beta 49 and 50, which bind to a site close to the N343 glycan). Analysis of the Omicron mAbs showed CDR3 sequences to be much longer, indicating a different binding mode to Beta 49, 50. In the Beta group of mAbs, the expansion of the public response was mediated through VH 4-39 (which binds to an epitope around the 501Y mutation) (6 of 27 mAbs), most of which lost activity against BA.1, and notably none of the current Omicron mAb group is encoded by VH4-39.

[0333] We found higher levels of somatic mutations in both the heavy and light chains compared to the early pandemic group of mAbs Omicron (mean values of 9.00, 6.00, VH and VL of the early pandemic group 4.55, 4.25, respectively) compared to the early pandemic group of antibodies. These results are consistent with the evolution of increased Omicron affinity via somatic mutations in vaccine-induced memory B cells.

[0334] Omicron mAb has broad neutralizing effect on VoC.

[0335] Neutralization assays were performed on Victoria and all variants of interest, including Alpha, Beta, Gamma, Delta, and Omicron BA.1, against a group of 28 potent mAbs. Figure 2A -C, Tables 13 to 16 and 18). The possible sources of all these antibodies from vaccine-induced memory B cells are obvious, since in almost all cases Victoria's FRNT50 titer was at the high end of all VoCs tested for each mAb ( Figure 2A -C (Tables 13 to 16 and 18). Among the five mAbs, BA.1 had FRNT50 titers <10 ng / ml, while mAbs Omi-3, 8, 12, 18, and 24 were the most potent, with FRNT50 titers of 9, 8, 4, 6, and 7 ng / ml, respectively, and FRNT90 titers of 67, 42, 20, 18, and 35 ng / ml, respectively.

[0336] The data provided in Tables 13, 14, and 16 include some IC50 data obtained using fake virus constructs. The data in Table 18 include IC50 results obtained entirely from real virus constructs.

[0337] Of the 28 antibodies, 17 showed cross-reactivity to all VoCs, with FRNT50 titers differing less than 10-fold among all viruses. Omi-06, 24, 30, 31, 34, and 41 showed reduced or absent activity against Delta (3 of which belonged to the VH1-69 family) and may possess epitopes influencing the L452R Delta mutation (Delta shares T478K with BA.1). Omi-09 and 32 antibodies were poorly responsive to Beta and Gamma and may be sensitive to E484K found in Beta and Gamma, but tolerant to E484A variations in Omicron (Omicron shares N501Y and K417N with Beta, while Gamma shares N501Y and K417T). Finally, although 129 anti-NTD mAbs were isolated, only one of them, Omi-41, showed an FRNT50 titer <100 ng / ml. Omi-41 showed neutralizing activity against Victoria, Alpha, Beta, and Gamma, but no activity against Delta, possibly due to the unique spectrum of NTD variations found in Delta.

[0338] Neutralization of BA.1 compared to BA.1.1, BA.2 and BA.3

[0339] Lentivirus-based reporter was constructed and pseudotyped with S gene sequences of Victoria, BA.1, BA.1.1, BA.2 and BA.3. Neutralization assays for Omicron mAbs were performed in Figure 2B , Tables 14 and 18, there were few differences in the neutralization of BA.1, BA.1.1, BA.2 and BA.3 for most antibodies. However, there were also some notable exceptions; for Omi-8, 29 and 32, BA.2 neutralization was reduced 38, 3 and 158-fold, respectively, compared to BA.1, while for Omi-6, 24, 34 and 35, BA.1.1 neutralization was reduced 40.9, 10.8, 7.8 and 6.6-fold, respectively, compared to BA.1, and for Omi-39 and 40, the neutralization was knocked out. Omi-mAbs neutralization of BA.3 was similar to that of BA.2, except for Omi-06 and Omi-36, where BA.3 neutralization titers were significantly lower than BA.1 or BA.2. For some reason, NTD-binding mAb Omi-41 did not neutralize Victoria in the pseudovirus system, but did neutralize live virus, as was also found for early pandemic mAb 159, which showed potent activity against live virus, but no activity against pseudovirus.

[0340] Pseudovirus neutralization curves for the mAb panel isolated from early pandemic cases and the mAb panel isolated from Beta cases are shown in Figure 4A , B and Table 15, in most cases, neutralization titers were similar for BA.1, BA.1.1 and BA.2, but there were some differences, mAbs 40, 278 and 318 neutralized BA.2 better than BA.1, while 222, Beta 22, 29, 54, 55 and 56 neutralized BA.1 better than BA.2, and Beta-53, which binds tightly to the N343 glycan, showed reduced neutralization of BA.1.1.

[0341] Neutralization of antibodies developed for clinical use.

[0342] Finally, Victoria, BA.1, BA.1.1, BA.2 and BA.3 strains were tested for neutralization by mAbs being developed for clinical use, where many differences were found Figure 2CInterestingly, the activity of known antibody A (REGN10987) partially recovers for BA.2, but is still 308-fold lower compared to Victoria, the activity of known antibody D (AZD1061) is almost fully recovered for BA.2, whereas known antibody D (AZD8895) is 5.4-fold lower for BA.2 compared to BA.1 and the combination of known antibodies D (AZD8895) and E is only 8-fold lower compared to Victoria. The activity of known antibody K (S309) is 6.8-fold lower for BA.2 compared to BA.1 and finally the activity of known antibody G (ADG20) is completely lost for BA.2.

[0343] In summary, the neutralization of most Omicron monoclonal antibodies is not affected by the differences between BA.1, BA.1.1, BA.2 or BA.3 mutations. However, some monoclonal antibodies do show differences, in particular known antibody A (REGN 10987) and known antibody C (AZD1 061) neutralize BA.2 more easily than BA.1 and known antibody K (S309) shows a decrease in neutralization for BA.2, which can prompt a sub-lineage classification before use. A structural explanation for the differences in neutralization between BA.1, BA.1.1, BA.2 and BA.3 will be discussed below.

[0344] Neutralization of BA.1, BA.1.1, BA.2 and BA3 by immune sera

[0345] To determine whether the difference in transmissibility between BA.1 and BA.2 can be due to different neutralization, but also to determine whether BA.2 can escape the BA.1 antibody response, neutralization assays were performed using sera from various sources. First, sera collected from vaccinees who received either the Oxford / AstraZeneca AZD1222 (n=41) or the Pfizer / BioNTech BNT162b2 (n=20) vaccine were used for neutralization assays against Victoria, BA.1, BA.1.1, BA.2 and BA.3 (Figure 6A, B). Figure 3 A, B).

[0346] For AZD1222, samples were collected 4 weeks after the second and third doses of vaccine. There was a small but significant difference in pseudovirus neutralisation after the third dose of AZD1222, with titres against BA.2 reduced relative to BA.1 (1.17-fold, p=0.0019) and titres against BA.1.1 reduced relative to BA.1 (1.29-fold, p=0.0086). For BNT162b2, samples were collected 4 weeks and 6 months after the second dose of vaccine, before the third dose and 4 weeks after the third dose. After the third vaccine dose, titres against BA.1, BA.2 and BA.1.1 were similar, with no significant difference between them. Next, the neutralisation of sera collected from cases infected with Omicron was determined. Early samples (n=12) were collected <14 days (median 13 days) after symptom onset and late samples (n=17) were collected >21 days (median 38 days) after symptom onset. All cases had received at least 2 doses of vaccine and many late convalescent cases received a third dose of vaccine after infection with Omicron. Neutralisation against Victoria, Alpha, Beta, Gamma, Delta and Omicron was tested using live virus neutralisation assays (Figure 1C). At early time points, all vaccinated cases had very high titres against Victoria, with a geometric mean FRNT50 close to 1 / 3000 and showed broad neutralisation against VoCs, with FRNT50 >1 / 1000 (FRNT50 = 558) for all viruses except Omicron. At late time points, titres against Victoria did not change, while titres against VoCs and Omicron increased (3-fold, p=0.0123). Pairwise comparisons of early and late samples taken from the same individual confirmed that the reaction to Omicron infection was broadly boosted (Figure 1A). Figure 3 C). At early time points, all vaccinated cases had very high titres against Victoria, with a geometric mean FRNT50 close to 1 / 3000 and showed broad neutralisation against VoCs, with FRNT50 >1 / 1000 (FRNT50 = 558) for all viruses except Omicron. At late time points, titres against Victoria did not change, while titres against VoCs and Omicron increased (3-fold, p=0.0123). Pairwise comparisons of early and late samples taken from the same individual confirmed that the reaction to Omicron infection was broadly boosted Figure 5 A).

[0347] Victoria, BA.1, BA.1.1, BA.2 and BA.3 neutralisation was determined by pseudovirus neutralisation. BA.1 neutralisation titres were higher at late time points. However, all sera were obtained from BA.1 infected cases and BA.2 neutralisation titres were reduced by a small but significant amount relative to BA.1 (1.7-fold and 1.5-fold at <14 days and >21 days respectively, p=0.0034 and 0.0067), BA.1.1 titres were not significantly reduced relative to BA.1, while titres against BA.3 were reduced by 1.7-fold relative to BA.1 at >21 days (p=0.0012) Figure 3 D, 5B).

[0348] In summary, good neutralizing titers of antibodies against Omicron BA.1 BA.1.1, BA.2, and BA.3 were induced after three doses of vaccine, particularly BNT162b2, with only minor differences between titers against BA.1 BA.1.1, BA.2, and BA.3. This can suggest that the increased transmissibility of BA.2 is not due to increased vaccine escape. After breakthrough Omicron infection, broad antibody responses to the variants of interest were boosted in previously vaccinated individuals, and a strong response to Omicron was generated. Since there was only a small difference in neutralization between BA.1 and BA.2, it is unlikely that BA.2 superinfection occurred in cases of BA.1 exposure and vaccination, at least in the short term.

[0349] Neutralization of BA.4 compared to BA.1, BA.1.1, BA.2, and BA.3

[0350] Neutralization of BA.4 / 5 compared to Omicron sublineages BA.1, BA.1.1, BA.2, BA.3, and the early pandemic Victoria strain was also assessed. BA.4 / 5 has been shown to have a more extreme antibody escape phenotype than BA.1 and BA.2, and neutralization titers of sera from triply vaccinated donors were reduced by about 2-3-fold compared to neutralization of BA.1 and BA.2. In addition, sera from vaccinees with breakthrough BA.1 infection showed a reduction in neutralization titers of about 2-3-fold against BA.4 / 5 compared to BA.1 and BA.2. This suggests that currently approved vaccines and mAbs can be less effective at preventing BA.4 / 5 transmission. Therefore, new monoclonal antibodies and combinations can be needed to fill the gap to protect the extremely vulnerable and those who cannot mount a sufficient vaccine response.

[0351] Neutralization of BA.4 by vaccine sera

[0352] A panel of pseudotyped lentiviruses expressing the S gene from Omicron sublineages BA.1, BA.1.1, BA.2, BA.3, and BA.4 / 5 (Di Genova et al., 2020, “Production, titration, neutralisation and storage of SARS-CoV-2 lentiviral pseudotypes”. Figshare preprint.) were constructed, as well as the early pandemic ancestral related strain Victoria, which was used as a control.

[0353] Sera obtained 28 days after a third dose of Oxford-AstraZeneca vaccine ADZ1222 (n=41) (Flaxman et al., 2021, “Reactogenicity and immunogenicity after a late second dose or a third dose of ChAdOxl nCoV-19 in the UK: a substudy of two randomised controlled trials (COV001 and COV002). Lancet 398, 981-990) or Pfizer-BioNTech vaccine BNT162b2 (Cele et al., 2021, “Omicron extensively but incompletely escapes Pfizer BNT162b2 neutralization”. Nature 602, 654-666) (n=20) Figure 7 A, B). For AZD1222, neutralisation titres were 2.1-fold lower for BA.4 compared to BA.1 (p=0.0001) and 1.8-fold lower for BA.2 (p=0.0001). For BNT162b2, neutralisation titres were 3.2-fold lower for BA.1 (p=0.0001) and 3.1-fold lower for BA.2 (p=0.0001) compared to the original Wuhan strain. These reductions in titres have the potential to reduce vaccine effectiveness, particularly at longer time points, as antibody titres naturally wane.

[0354] Neutralisation of BA.4 / 5 by sera from breakthrough BA.1 infection

[0355] At the start of the Omicron surge, vaccinated volunteers who had experienced a breakthrough Omicron infection were recruited. Samples were first taken at <14 days (median 13 days) after symptom onset, while later samples were taken at >21 days (median 38 days) after symptom onset (n=16). Pseudovirus neutralisation assays were performed on pseudovirus panels representing variants of concern and Omicron sublineages Figure 7 C, D).

[0356] BA.1 infection after vaccination leads to a broad neutralization response with high titers against all VoCs, which is boosted at later time points (Nutalai et al., 2022, “Potent cross-reactive antibodies following Omicron breakthrough in vaccines”. Cell (in press)). Neutralization titers against BA.4 were significantly lower than against BA.1 and BA.2, and at early time points, BA.4 / 5 titers were 1.9-fold (p=0.0001) and 1.5-fold (p=0.0015) lower than BA.1 and BA.2, respectively. At later time points, BA.4 / 5 titers were 3.4-fold (p=0.0001) and 2-fold (p=0.0017) lower than BA.1 and BA.2, respectively.

[0357] Thus, BA.4 / 5 show some degree of immune evasion of the vaccine / BA.1 response when compared to BA.1 and BA.2. These samples were all collected quite close to the time of infection, which means that further attenuation over the intervening months can have made the individuals susceptible to reinfection with BA.4 / 5.

[0358] BA.4 / 5 escape monoclonal antibodies

[0359] Sensitivity to L452R: It has been previously reported that Omi-24, 30, 31, 34 and 41 show complete knock-out of neutralization activity against Delta, while Omi-06 shows a severe knock-down of activity (Nutalai et al., 2022). Since BA.1 and BA.2 have only one of the two Delta RBD mutations (T478K), and BA.4 / 5 also have L452R, it was expected that all five L452-directed mAbs would be knocked out on BA.4 / 5. This was indeed observed (Table 20, A). Omi-41 was also unable to neutralize, which was attributed to a difference in the mutation in the NTD (A). Figure 8 Figure 9 A).

[0360] To confirm that the observed neutralization was directly attributable to changes in RBD interaction, the binding of selected antibodies to BA.4 / 5 and BA.2 RBD was also analyzed by surface plasmon resonance (SPR) (B, C). Omi-31 was selected as a representative of the L452R-sensitive antibody group, and as expected, binding was severely affected (B, C). Figure 10 、 15 ) (B, C). Figure 10

[0361] ​​As detailed information on the interaction of several Omicron-reactive antibodies with RBD was available, the BA.4 / 5 RBD mutations were modelled in the context of the known structure of the Omicron Fabs complexed with BA.1 or Delta RBD (Dejnirattisai et al., 2022, “SARS-CoV-2 Omicron-B.1.1.529 leads to widespread escape from neutralizing antibody responses”. Cell 185, 467-484 e415; Nutalai et al., 2022) Figure 11 ). The Omi-31 complex is shown in Figure 11 A and shows that L452 is neatly jammed into a hydrophobic pocket that cannot accommodate the larger positively charged arginine in BA.4 / 5 and Delta.

[0362] L452R enhances binding: Omi-32 neutralises BA.4 / 5 77-fold better than BA.2. Kinetic analysis of Fab binding to RBD shows that this is achieved mainly by a 5-fold increase in the on-rate Figure 10 B, C). This is explained mainly by a favourable salt bridge between the arginine at 452 and residue 99 of the heavy chain (HC) CDR3 Figure 11 B), perhaps assisted by removal of a slightly unfavourable charge interaction at residue 493. These electrostatic changes have the potential to increase the on-rate by electrostatic steering of the incoming antibody.

[0363] Sensitivity to F486V: Extending the logic used to understand Delta sensitivity, the remaining antibodies active against BA.4 / 5 > BA.2 but against Delta are likely to be sensitive to the F486V change, namely Omi-02, 09, 12, 23, 25, 26 and 29. Binding sensitivity was confirmed by SPR analysis on Omi-12 Figure 10 D, E), showing almost a 1,000-fold decrease in affinity. The Omi-25 complex provides an example of the structural basis for sensitivity Figure 11 C), showing the phenylalanine side chain acting as a binding hotspot, nestled in a hydrophobic cavity, making a favourable ring stacking interaction with Y106 of the HC CDR3.

[0364] Activity of commercial antibodies against BA.4 and BA.5

[0365] A panel of antibodies developed for therapeutic / prophylactic use was tested against BA.4 / 5 (Figure 8 B, Table 21). Many of these antibodies have had their activity against BA.1, BA.1.1 or BA.2 severely reduced or knocked out. For AstraZeneca AZD1061, activity against BA.4 / 5 was similar to BA.2 (reduction <2-fold), while for AZD8895, residual activity against BA.2 was knocked out. Activity of the combination of two antibodies in AZD7442 (Dong et al., 2021, “Genetic and structural basis for recognition of SARS-CoV-2 spike protein by a two-antibody cocktail”. Nature Microbiol. 6, 1233-1244) was reduced 8.1-fold compared to BA.2. Residual activity of REG10987 against BA.2 (Weinreich et al., 2021, “REGN-COV2, a Neutralizing Antibody Cocktail, in Outpatients with Covid-19”. N Engl J Med 384, 238-251) was further reduced on BA.4 / 5, as was residual BA.1 neutralization activity against ADG20 on BA.4 / 5 (Yuan et al., 2022, “A broad and potent neutralization epitope in SARS-related coronaviruses”. bioRxiv. https: / / doi.org / 10.1101 / 2022.03.13.484037). For S309 (VIR-7831 / 7832) (Sun and Ho, 2020, “Emerging antibody-based therapeutics against SARS-CoV-2 during the global pandemic”. Antib Ther 3, 246-256), activity against BA.4 / 5 was reduced 1.6-fold compared to BA.2.

[0366] These effects can be rationalized in terms of the way the reference antibody interacts with the RBD, for example in the case of AZD8895 (an IGHV1-58 genotype mAb, Figure 11 E) F486 forms a hydrophobic interaction hotspot that would be abolished by mutation to a much smaller valine side chain. Antibody residues involved in interaction with F486 are highly conserved in mAbs of this genotype (including Omi-12, 253 and Beta-47) (Nutalai et al., 2022, “Potent cross-reactive antibodies following Omicron breakthrough in vaccines”. Cell (in press); Dejnirattisai et al., 2021, “The antigenic anatomy of SARS-CoV-2 receptor binding domain”. Cell 184, 2183-2200 e2122; Liu et al., 2021, “The Beta mAb response underscores the antigenic distance to other SARS-CoV-2 variants”. Cell, Host and Microbe 30, 53-68), explaining the severe impact of the F486V mutation on neutralization of these mAbs Figure 8 A, 13).

[0367] Neutralization of BA.2.75 by vaccine sera

[0368] A panel of pseudotyped lentiviruses expressing the S gene from Omicron sublineages BA.1, BA.1.1, BA.2, BA.2.12.1, BA.4 / 5, BA.2.75 was constructed as above (Di Genova et al., 2020), as well as Victoria (an early pandemic ancestral related strain) used as a control. D339H, G446S, N460K and R493Q were also included as single mutations in the BA.2 background. Neutralisation assays were performed with sera obtained 28 days after a third dose of Oxford-AstraZeneca vaccine AZD1222 (n=41) (Flaxman et al., 2021 “Reactogenicity and immunogenicity after a late second dose or a third dose of ChAdOxl nCoV-19 in the UK: a substudy of two randomised controlled trials (COV001 and COV002)”. Lancet 398, 981-990) or Pfizer-BioNTech vaccine BNT162b2 (n=22) (Cele et al., 2021; “Omicron extensively but incompletely escapes Pfizer BNT162b2 neutralization”. Nature 602, 654-666e) Figure 17 ) For AZD1222, neutralisation of BA.2.75 was 1.2-fold lower than for BA.2 (p=0.0182) and 1.1-fold lower than for BA.2.12.1 (p=0.0065) but 1.5-fold higher than for BA.4 / 5 (p<0.0001) Figure 17 B). Overall, there was a reduction in the neutralisation titre of BA.2.75 in vaccine sera compared to BA.2, but not to the level seen for BA.4 / 5.

[0369] Neutralisation of BA.2.75 by sera from vaccine breakthrough BA.1 or BA.2 infection

[0370] Breakthrough BA.1 serum samples were collected from vaccinated volunteers >28 days from symptom onset (median 38 days; n=16). Pseudovirus neutralisation assays were performed on the pseudovirus panel described above Figure 17C). Neutralization titers for BA.2.75 were similar to BA.2, 1.4-fold higher than BA.2.12.1 (p=0.0052) and 2.0-fold higher than BA.4 / 5 (p=0.0001), suggesting that BA.2.75 can be less likely than BA.2.12.1 or BA.4 / 5 to cause reinfection in individuals who have experienced breakthrough infection with BA.1.

[0371] Breakthrough BA.2 serum samples were collected from vaccinated volunteers >12 days after symptom onset (median 29 days; n=23). Pseudovirus neutralization assays were performed on pseudovirus panels Victoria, BA.1, BA.1.1, BA.2, BA.2.12.1, BA.4 / 5, and BA.2.75 Figure 17 D). Here, neutralization titers for BA.2.75 were significantly reduced compared to BA.2 (1.4-fold; P=0.0021), similar to BA.2.12.1, but still higher than BA.4 / 5 (1.4-fold; P=0.0123). Taken together, BA.2.75 shows some degree of escape from the humoral response induced by breakthrough infection with BA.2, but not BA.1.

[0372] Individual BA.2.75 mutations have different effects on neutralization

[0373] To understand the impact of individual mutations in the BA.2.75 RBD, these mutations were introduced individually into the pseudovirus BA.2 background and their neutralization was assayed using Pfizer BNT162b2 serum from a trivalent vaccinee Figure 17 E). Of the 4 single mutation variants of BA.2, 3 reduced neutralization titers for BA.2, with N460K reducing the most (3.1-fold, p<0.0001), followed by D339H (1.3-fold, p=0.0006), and then G446S (1.2-fold, p=0.2312), whereas the R493Q back mutation increased the neutralization titer by 1.5-fold (p<0.0001). Q493 is present in all vaccines, explaining the increased activity of vaccine serum on this back mutation.

[0374] BA.2.75 escapes monoclonal antibodies

[0375] To dissect how BA.2.75 can affect neutralizing antibody activity, a panel of potent human mAbs recently reported to be generated from cases of Omicron breakthrough infection (BA.1 IC50 titers <0.1 pg / ml) were tested using pseudovirus assays (Nutalai et al., 2022) Figure 19A, Table 22). Of the 27 RBD-specific mAbs, those belonging to the IGHV3-53 / 66 family were most severely affected. Three (Omi-16, Omi-29 and Omi-36) showed complete knock-out of BA.2.75 neutralisation; another four (Omi-18, Omi-20, Omi-27 and Omi-28) showed >5-fold reduction compared to BA.2, consistent with the observation that the GGS / T motif of CDR-H2 interacts highly conserved in the RBD / IGHV-3 / 66 complex structure (N460) Figure 21 B) (Dejnirattisai et al., 2021, Liu et al., 2021, Nutalai et al., 2022).

[0376] As with BA.2 and BA.4 / 5, BA.2.75 was not neutralised by anti-NTD mAb Omi-41, which only interacts with the NTD of BA.1, BA.1.1 and BA03.

[0377] The Omi mAbs were also tested against the above-mentioned pseudoviruses encoding single point mutations in the BA.2 RBD Figure 23 , Table 24). The VH3-53 / 66 mAbs that lost neutralisation of BA.2.75 were also affected by the N460K mutation, confirming the prediction that this residue is critical for the binding of many members of this public gene family. Interestingly, the BA.2 + N460K mutation alone showed greater impact on the activity of several mAbs than BA.2.75: the neutralisation titre of Omi-03 (IGHV3-53) was reduced 50-fold for BA.2 + N460K but only 2-fold for BA.2.75; Omi-17 (IGHV3-66) was completely knocked out on BA.2 + N460K but only reduced 4-fold for BA.2.75; and Omi-33 (IGHV3-33) was reduced 7-fold for BA.2 + N460K but no change was observed for BA.2.75. Thus, the other mutations of BA.2.75 can have mitigated the impact of the N460K mutation, particularly the R4930 mutation.

[0378] Interestingly, BA.2.75 was more sensitive to Omi-32 (IGHV-3-33) than BA.2, with an 8-fold increase in neutralisation titre. The Omi-32 enhanced activity can be due to the stronger interaction of the antibody with the RBD by the G446S mutation Figure 19 A, Table 22).

[0379] To confirm that the observed changes in neutralisation activity were related to changes in RBD interactions, the binding of selected antibodies to BA.2.75 and BA.2 RBD was analysed by surface plasmon resonance (SPR) Figure 24). Omi-29 (IGHV3-53) and Omi-36 (IGHV3-66) were severely impaired in binding to BA.2.75 and Omi-18 and Omi-20 showed an 8-fold reduction in binding affinity to BA.2.75 compared to BA.2. On the other hand, a 2-fold increase in binding affinity of Omi-32 to BA.2.75 compared to BA.2 was observed, which is in line with the observed increase in neutralization titers.

[0380] Escape from commercial monoclonal antibodies against BA.2.75

[0381] A panel of mAbs (Table 23) developed as therapeutic agents against BA.2.75 was evaluated for their sensitivity. Figure 19 B, Table 23) were evaluated for their sensitivity. The neutralization profile between BA.2.75 and BA.2 was overall similar; however, in addition to 6 out of 12 mAbs that had completely lost neutralization activity against BA.2 (REGN10933, ADG10, ADG20, ADG30, Ly-CoV555, Ly-CoV16), REG10987 had residual activity against BA.2 (Weinreich et al., 2021, “REGN-COV2, a Neutralizing Antibody Cocktail, in Outpatients with Covid-19.” N Engl J Med [New Engl J Med]384, 238-251) were further knocked out against BA.2.75 due to the G446S mutation (Dejnirattisai et al., 2022). For AstraZeneca AZD1061, activity against BA.2.75 was similar to that against BA.2 (reduction <3-fold); whereas AZD8895 titres recovered from 1.333 pg / ml for BA.2 to 0.008 pg / ml for BA.2.75, with a 167-fold increase in activity. Therefore, AZD7442 (a combination of AZD8895 and AZD1061) (Dong et al., 2021, “Genetic and structural basis for recognition of SARS-CoV-2 spike protein by a two-antibody cocktail”. Nature Microbiol. 6, 1233-1244) showed similar activity against BA.2.75 and BA.2 (2-fold reduction). These results can be explained by the structure of the ternary complex of ancestral SARS-CoV-2 RBD / AZD1061 / AZD8895 (Dong et al., 2021). G446 is in contact with CDR-L2 Y55 and W56 of AZD1061, and the G446S mutation would cause steric clashes Figure 21 D, E). While CDR-H2 of AZD8895 sits above and forms a hydrogen bond with Q493 of the RBD, the arginine at 493 would cause a severe clash with CDR-H2 of the mAb Figure 21F, G). The S309 activity of BA.2.75 was increased 3-fold compared to BA.2 (Sun and Ho, 2020, “Emerging antibody-based therapeutics against SARS-CoV-2 during the global pandemic.” Antib Ther 3, 246-256), indicating that the D339H mutation in BA.2.75 reduced the impact of the previous G339D mutation in BA.2 on S309 activity. LY-CoV1404 (bebtelovimab) (Westendorf et al., 2022, “LY-CoV1404 (bebtelovimab) potently neutralizes SARS-CoV-2 variants.” Cell Rep 39, 110812) is the only mAb that is completely preserved in neutralizing all Omicron sublineages.

[0382] BA.2, BA.4, and BA.5 sublineage escape monoclonal antibodies

[0383] The decline in mAb activity was also observed in the new BA.2, BA.4, and BA.5 sublineages (including BA.4.6, BA.2.75, BA.2.75.2, BA.2.3.20, BJ.1, BQ.1, BQ.1.1, XBB, XBB.1, and XBB.1.5) Figure 34 ), with XBB producing the most extreme escape. The activity of all 9 IGHV3-53 / 66 mAbs was reduced >100-fold, with 5 / 9 being completely knocked out by BA.2.75.2. Only one mAb, Omi-42, was not affected by all variants. Omi-42 is unusual because it binds at the back of the RBD left shoulder (Nuta1ai et al., 2022), a region that has not been targeted for mutation by this set of newly emerging BA.2 variants, perhaps because antibodies that bind in this region are relatively rare.

[0384] Further data can be found in Nutalai et al. (2022) “Potent cross-reactive antibodies following Omicron breakthrough in vaccinees”, Cell 185(12), 2116-2131; Huo et al. (2022) “Humoral responses against SARS-CoV-2 Omicron BA.2.11, BA.2.12.1 and BA.2.13 from vaccine and BA.1 serum”, Cell discovery 8, 119; and Huo et al. (2022) “A delicate balance between antibody evasion and ACE2 affinity for Omicron BA.2.75”, Cell Reports, 42(1). 2023.

[0385] Neutralization of BA.2 subvariants BA.2.11, BA.2.12 and BA.2.13 by vaccine sera

[0386] The receptor binding ability of BA.2 subvariants BA.2.11, BA.2.12 and BA.2.13 was also evaluated. High resolution crystal structures of the BA.2.12.1 RBD were generated, showing different sensitivities of the new BA.2 subvariants BA.2.11, BA.2.12 and BA.2.13 compared to BA.2 to serum samples and monoclonal antibodies (mAbs).

[0387] Given the physicochemical properties of the side chain of residue 452, BA.2.13 is expected to be a relatively modest change; the change of L to M will increase the size of the side chain, but it will still be hydrophobic. The change of L to Q in BA.2.12.1 will introduce some polar character, while BA.2.11 is the most radical, with the change of L to R introducing a large basic amino acid.

[0388] Neutralization of BA.2 subvariants BA.2.11, BA.2.12 and BA.2.13 by vaccine sera

[0389] To assess the sensitivity of BA.2 subvariants to neutralization by immune sera, a series of serum samples were used to perform neutralization assays on pseudotyped lentiviruses expressing the spike gene of BA.2.11, BA.2.12 and BA.2.13.

[0390] First, sera collected 4 weeks after the third dose of the Oxford-AstraZeneca vaccine AZD1222 (n=41) or the Pfizer-BioNTech vaccine BNT162b2 (n=18) were observed for neutralization. No significant loss of neutralization titers was seen compared to BA.2. In fact, BA.2.13 showed a significant increase in AZD1222 vaccinees (1.6-fold, p<0.0001) Figure 27 a, b). This contrasts with a recent report (Cao, Y. et al. “BA.2.12.1, BA.4 and BA.5 escape antibodies elicited by Omicron infection”. Nature, 2022) where sera collected from three-dose vaccinees (four weeks after the third dose of the inactivated vaccine CoronaVac or a booster dose of ZF2001 after two doses of CoronaVac) showed a significant reduction in neutralization titers for both BA.2.12.1 and BA.2.13 (BA.2.11 was not tested).

[0391] Neutralization of BA.2 subvariants BA.2.11, BA.2.12 and BA.2.13 by sera from vaccine breakthrough BA.1 or BA.2 infections

[0392] Next, sera samples collected from vaccinees infected with BA.1 were examined for neutralization. The samples (n=14) were taken >28 days (median 38 days) after the onset of symptoms; all convalescent individuals had received at least 2 doses of vaccine, of which 3 received a third dose of vaccine after infection with Omicron. There was a significant reduction in neutralization titers for all three variants compared to BA.2, with the greatest reduction for BA.2.11 (1.6-fold, P=0.0067), followed by BA.2.12.1 (1.4-fold, P=0.0085) and BA.2.13 (1.2-fold, P=0.0085) Figure 27c). In summary, these observations suggest that the subvariants do not show stronger humoral immune evasion in individuals vaccinated with three doses of AZD1222 or BNT162b2 vaccine compared to BA.2. However, for vaccinees who suffered a BA.1 breakthrough infection, regardless of the type of vaccine they had received, the BA.2 variant was more able to evade the humoral response, despite the induction of a broad neutralizing antibody response with high titres against all the variants of interest (Nutalai, R. et al., “Potent cross-reactive antibodies following Omicron breakthrough in vaccines [Omicron breakthrough in vaccines following potent cross-reactive antibodies]”. Cell, 2022. 185(12): p. 2116-2131, e18). This can indicate a different selection pressure for BA.2 and its subvariants in the high background of breakthrough infections. Since antibody titres naturally wane at longer time points, it is expected that people who suffered a BA.1 breakthrough infection are more susceptible to reinfection with BA.2 subvariants.

[0393] To further elucidate the different responses between BA.2 and its subvariants, a panel of potent human monoclonal antibodies (mAbs) raised from BA.1 breakthrough infection cases were subjected to pseudovirus assays (Nutalai et al., 2022) Figure 29 ). Consistent with the structural observations and neutralization results, the most reduction in neutralization titres was observed for BA.2.11, followed by BA.2.12.1 and BA.2.13, with complete knock-out of neutralization for BA.2.11 for 5 out of 27 mAbs (Omi-06, Omi-24, Omi-30, Omi-31 and Omi-34). There was also a varying degree of reduction in neutralization activity of the same panel of mAbs against BA.2.12.1, while there was little change for BA.2.13. Among them, Omi-06 belongs to the IGVH4-4 family, while the other four mAbs belong to the IGVH1-69 family. In fact, previous structural studies predicted that Omi-06 and Omi-31 are sensitive to the L452R mutation in Delta (Nutalai et al., 2022). To confirm that the observed different neutralization is directly attributable to changes in RBD binding, surface plasmon resonance (SPR) was used to compare the binding behaviour of BA.2 and BA.2.12.1 RBDs, using Omi-06 and Omi-31 as examples. As expected, the affinity was reduced, with 15-fold weaker binding of BA.2.12.1 RBD than BA.2 for Omi-06, and strikingly, about 1300-fold weaker binding of BA.2.12.1 RBD to Omi-31 Figure 30

[0394] ​Spike mutations in the BA.2.11, BA.2.12, and BA.2.13 variants may make them slightly more transmissible than BA.2. However, compared to BA.2, they did not appear to acquire greater humoral immune escape in healthy vaccine recipients who had received three doses of the Oxford-AstraZeneca or Pfizer-BioNTech BNT162b2 vaccine. This result differs from that of vaccine recipients who had received three doses of the CoronaVac vaccine, who had significantly reduced neutralizing titers (Cao, Y. et al., BA.2.12.1, BA.4 and BA.5 escape antibodies elicited by Omicron infection. Nature, 2022). However, in vaccine recipients who experienced a breakthrough infection with BA.1, regardless of the type of vaccine received, a significant decrease in neutralizing titers was observed. This may be partly due to the partial or complete knockout of neutralizing activity of antibodies belonging to the IGVH1-69 family, many of which are sensitive to mutations at leucine 452 of the spike RBD. This suggests that the evolving Omicron sublineage can evade humoral immune responses induced by BA.1, implying that the BA.1 spike virus or RBD may not be a better immunogen than its ancestral strains in developing next-generation SARS-CoV-2 vaccines.

[0395] BA.2.12.1 Crystal Structure

[0396] The crystal structure of BA.2.12.1RBD was determined at 2.38 Å to be a ternary complex with neutralizing Fab and nanobody. Figure 27 The result (hm) indicates that the structural differences are essentially limited to the side chain of residue 452.

[0397] The neutralization of BA.2.75.2 by mAb prepared after BA.1 infection.

[0398] A panel of mAbs prepared after BA.1 infection were investigated for their neutralization of BA.2.75.2 (Nutalai et al., 2022). A decrease in mAb activity against BA.2.75.2 was observed (Table 32a). The activity of all nine IGVH3-53 / 66 mAbs was reduced by >100-fold, with four out of nine showing complete knockout of BA.2.75.2 activity. Only one mAb, Omi-42, was unaffected by all variants and showed neutralizing activity against the BA.4+14 mutation, with an IC50 of 11 ng / ml. Omi-42 is unusual because it binds to the posterior part of the left shoulder of the RBD (Nutalai et al., 2022), a region that has not yet been targeted for mutation, perhaps because antibodies binding to this region are relatively rare.

[0399] A set of mAbs developed for clinical use was also tested (Dong et al., 2021; Sun and Ho, 2020; Weinreich et al., 2021; Yuan et al., 2022). Many of these mAbs were severely impacted by many variants. The activity of all mAbs except S309 was knocked out by one or more variants, including Ly-CoV1404 (Westendorf et al., 2022) (see Table 32b).

[0400] Pfizer BNT162b2 vaccine serum neutralization titers for BA.2.75.2 and BA.2.3.20.

[0401] Neutralization of sera collected 28 days after the third dose of Pfizer BNT162b2 vaccine (Polack et al., 2020) and characteristics of these subjects in cases of infection with BA.1, BA.2, or BA.4 / 5 are described in the Methods.

[0402] Using sera obtained 28 days after BNT162b, the infectious titers of BA.2.75.2 showed a large reduction compared to BA.2 and BA.4 and were the lowest of all tested variants compared to the ancestral strain Victoria. The titers of BA.2.75.2 were reduced compared to BA.2.75, which showed only a modest reduction compared to BA.2. The titers of BA.2.3.20 were also reduced substantially.

[0403] Using sera obtained after infection with BA.1, BA.2, and BA.4 / 5, the neutralization titers of BA.2.75.2 and BA.2.3.20 were similarly reduced substantially compared to BA.4 / 5. Neutralization of the BA.4 + 14 RBD mutations described above was also reduced compared to BA.2 and BA.4 / 5, but not much more than BA.2.75.2, indicating a dominant effect of the mutations in BA.2.75.2.

[0404] Neutralization of BA4.6 by sera from vaccine breakthrough BA.1 or BA.2 infection

[0405] Here, we use the following to investigate the neutralization of BA.4.6: Pfizer-BioNTech vaccine sera, BA.1, BA.2, and BA.4 / 5 vaccine breakthrough immune sera, and panels of monoclonal antibodies. Notably, we show further antibody evasion of BA.4.6, providing guidance for vaccine design and the use of therapeutic monoclonal drugs.

[0406] To assess the antibody evasion capacity of BA.4.6, we constructed a panel of pseudotyped lentiviruses expressing S genes from BA.4.6 and other SARS-CoV-2 variants (Di Genova, C. et al. Production, titration, neutralisation and storage of SARS-CoV-2 lentiviral pseudotypes. figshare, 2020), and the early pandemic ancestral related strain Victoria, used as a control. First, sera collected 4 weeks after the third dose of the Pfizer-BioNTech vaccine BNT162b2 (n=22) were examined for neutralisation. BNT162b2 sera had a 2-fold reduction in neutralisation titre against BA.4.6 compared to BA.4 / 5 (p<0.0001) Figure 28 a).

[0407] Sera samples collected from vaccinees infected with BA.1 were assayed for neutralisation. Samples (n=16) were taken >28 days after symptom onset. BA.2 or BA.4 / 5 samples (n=23) were taken >12 days after symptom onset. Samples (n=11; all vaccinated except one) were taken >23 days after symptom onset Figure 28 b-d). There was a significant reduction in neutralisation titre of breakthrough BA.1 (1.5-fold; P=0.0006) and BA.2 (1.2-fold; P=0.0384) sera samples against BA.4.6 compared to BA.4 / 5. Notably, BA.4.6 was able to efficiently escape neutralisation from sera samples from BA.1 breakthrough infections, showing a large reduction in titre compared to BA.1 (4.4-fold; p=0.0001), BA.2 (3-fold; p=0.0009) and BA.4 / 5 (1.5-fold; p=0.0006). A small, non-significant increase in neutralisation titre against BA.4.6 was observed in the BA.4 / 5 breakthrough cohort compared to BA.4 / 5.

[0408] To further characterise the antigenic escape properties of BA.4.6, a panel of potent human mAbs raised from BA.1 breakthrough convalescents were assayed on pseudotypes (Nutalai et al. 2022) Figure 28e) In summary, the neutralization profile of BA.4.6 was similar to BA.4 / 5. However, the residual activity of Omi-35 (IC50= 1.687 pg / mL) was further knocked out against BA.4.6, and the potency of Omi-32 and Omi-33 against BA.4 / 5 (IC50= 0.035 and 0.013 pg / mL, respectively) was completely impaired against BA.4.6. The loss of activity of Omi-32 can be explained by the disruption of the interaction between H1 and R346, as illustrated by previous structural analysis (Nutalai et al., 2022).

[0409] Neutralization of BA.4.6 by clinically used mAbs

[0410] Finally, the neutralizing activity of a number of clinically used mAbs was evaluated Figure 28 f) The potency of AZ1061 / cilgavimab against BA.4 / 5 was completely knocked out against BA.4.6, resulting in the complete loss of activity of AZ7742 / Evusheld (a combination of AZ1061 / cilgavimab and AZ8895 / tixagevimab, which already had no activity against BA.4 / 5). The activity of S309 / sotrovimab was further reduced compared to BA.2 and BA.4 / 5 (not authorized for COVID-19 treatment by the US Food and Drug Administration, FDA, since April 2022 due to no efficacy against BA.2). Therefore, this leaves Ly-Covl404 / betulovimab as the only option to treat BA.4.6.

[0411] In summary, BA.4.6 showed further reduced neutralization from sera of three-dose Pfizer vaccinees as well as BA.1 and BA.2 vaccine breakthroughs compared to BA.4 / 5. Notably, BA.4.6 did not appear to be more resistant to neutralization from sera of BA.4 / 5 breakthrough infections compared to other variants. This collectively indicates that unless one has been triple-vaccinated and recovered from a BA.4 / 5 infection, it is highly likely to be infected or breakthrough infected with BA.4.6, which appears to offer some protection against BA.4.6.

[0412] As of September 2022, bivalent booster vaccinations combining ancestral strains with Omicron BA.1 are being rolled out in the UK and have recently been authorized by the FDA. The effectiveness of these bivalent boosters in preventing BA.4.6 infections remains to be seen. Finally, BA.4.6 further impaired the activity of EvuSheld, which remained active against BA.4 / 5; therefore, now only LY-CoVl404 / betulovimab retains potency against all circulating SARS-CoV-2 variants.

[0413] Systematic theme of mAb interactions

[0414] Omi-3 (a representative of the IGVH3-53 gene family) and AZD8895 (IGVH1-58) both have an association with F486. While the F486V mutation has little effect on Omi-3 (Table 2), Figure 10 F, G, 11F), it severely reduces the neutralization of AZD8895 and other IGVH1-58 mAbs (e.g., Omi-12) (Table 2, Figure 10 D, E, 11E). Notably, while many of the Omi series antibodies belonging to the closely related IGVH3-53 and IGVH3-66 gene families (9 of 28, Figure 8 A, Table 21) are almost perfectly adapted to the changes in BA.4 / 5, most of the antibodies from these gene families that were raised against earlier variants are knocked out on BA.1 and BA.2 (Nutalai et al., 2022), which is consistent with selection by breakthrough Omicron infection for a subset of antibodies that are insensitive to further BA.4 / 5 mutations.

[0415] The impact can vary greatly for antibodies with broadly similar epitopes, which is also true for antibodies with a 452 or 486 center in their binding footprint. Thus, Omi-31 (IGVH1-69) and Omi-32 (IGVH3-33) both bind in the front of the right shoulder, with the position of their CDR-H3 close to 452, while the activity of Omi-31 is abolished by L452R (as described above) and that of Omi-32 is significantly enhanced (Table 2, Figure 8 A, 11A, B). Similarly, Omi-25 and Omi-42 both belong to the IGVH3-9 gene family, and their footprints are both in the 486 region (Table 2, Figure 11 C, D). Omi-25 contacts F486, so the neutralization of BA.4 / 5 is abolished. In contrast, Omi-42 does not contact either of the mutated sites, and the neutralization of BA.4 / 5 is fully preserved (Table 2, Figure 10 H, I, 11D).

[0416] RBD antibody binding is finely mapped using competition measurements.

[0417] A matrix of pairwise BLI measurements is used to map potent RBD-binding Omicron mAbs and several pre-pandemic mAbs with known binding positions.

[0418] This approach yielded consistent predictions. The mAbs segregated into a limited set of epitopes that appeared to be a subset of the epitopes observed for the earlier pandemic viruses and were distinct from the lesions observed for Beta. Essentially, the antibodies clustered in two regions, one including VH3-53 and VH3-66 type antibodies, toward the back of the neck / left shoulder, extending all the way to the top of the left shoulder, while the other was in the front of the right shoulder region of the neck, toward the S309 known antibody binding site. This region was occupied by VH1-69 family antibodies, except for Omi-2, which was in the other cluster. The position of mAb Omi-09, which showed reduced neutralization of Beta and Gamma, was close to residue 484, which is mutated from Glu to Lys in Beta / Gamma and to Ala in Omicron. The VH1-69 mAbs Omi-24, 30, 31, and 34, which showed reduced neutralization of Delta, were close to residue 452, which is mutated from Leu to Arg in Delta.

[0419] Structure of anti-Omicron Fab / RBD complexes

[0420] Structural analysis of selected potent Omicron mAbs. The crystal structure of the Omicron BA.1 RBD in complex with 3 different Fabs was determined: Omi-3, 9, and 12. The complex of Omi-12 was at low resolution The structure of the individual Fabs was therefore determined at high resolution and rigid body fitted to obtain the complex structure.

[0421] Omi-3 belongs to the VH3-53 gene family and demonstrates how this gene family is adapted to broadly neutralize all major SARS-CoV-2 variants (however, like all potent Omicron antibodies, it does not bind the SARS-CoV-1 RBD). One fundamental problem with these antibodies is that most VoCs have the N501Y mutation, which introduces a steric clash with the LC CDR1 (L1) that can abrogate binding of most VH3-53-containing antibodies. However, two mechanisms of displacing L1 to avoid this clash have been reported previously (Dejnirattisai et al., 2021b; Liu et al., 2021b). In mAb-222, isolated from an individual infected early in the pandemic with a strain of the Wuhan outbreak, a proline was inserted at residue 30, which can jam Tyr-501 without a clash (Dejnirattisai et al., 2021b), thus enabling it to neutralize the Alpha, Beta, and Gamma variants. Beta-27 uses an alternative mechanism, lengthening the HC CDR3 (H3) loop from the usual 9 residues to 11 residues, displacing L1 to create sufficient space to allow 501Y to be stabilized by main chain interactions, thus conferring similar cross-reactivity (Liu et al., 2021b).

[0422] Omi-3 uses the same mechanism as Beta-27 to accommodate the N501Y mutation, although the Omi-3 H3 is one residue longer. The other VH3-53 Omicron antibodies (Omi-18 and Omi-29) have very similar H3s to Beta-27 and likely use the same mechanism. This L1 configuration is also compatible with the Y505H mutation in Omicron. However, neither 222 nor Beta-27 can effectively neutralize Omicron, which is likely due to the specific features of the H3 loop being tightly linked to the Q493R Omicron mutation.

[0423] Omi-9 is one of three VH3-30 mAbs and binds to the left shoulder of the RBD. Omi-9 shows relatively weak neutralization of Beta and Gamma (Figure 2). Other antibodies with high sequence similarity bind similarly, with the H3 contacting residue 484. Although the resolution of the Omi-9 / BA.1 complex is lower But it is clear that the H3 contacts residue 484, explaining the sensitivity to E484K in Beta and Gamma, while E484A in Omicron is tolerable.

[0424] Omi-12 belongs to the VH1-58 gene family (it is the only member of this family among the 28 potent Omicron antibodies). Like Omi-12, several members of this gene family have a glycosylation site at residue 102 of the heavy chain CDR3, whose role is unclear. VHl-58 antibodies raised during the early pandemic or Beta virus infection show reduced ability to neutralize Omicron, for example mAb 253, Beta-47 and known antibody D (AZD8895) show reduced activity of Omicron BA.1 relative to Victoria, respectively.

[0425] In contrast, Omi-12 has adapted and can efficiently neutralize Omicron and all VoCs Figure 2A , B) VH1-58 antibodies bind to the left shoulder epitope, H3 contacts S477N, but mutations at this position in Iota have no effect on VH1-58 mAb neutralization using pseudovirus assays. In addition, mAb 253 is able to neutralize Delta despite the T478K mutation. BA.2 with early pandemic mAb 150 (VH3-53). Detectable residual activity is observed in BA.1, BA.1.1 and BA.2 (BA.3 not tested). Two complex structures were obtained in different space groups, which are very similar and provide 3 independent views of the complex. mAb 150 binds in a similar pose to that observed previously for early pandemic viruses, but it is translated and forms looser interactions, which is consistent with almost complete loss of neutralization activity. This shows the huge impact of the adaptation mutations found in Omi-3.

[0426] Interestingly, in BA.2, three of the serine residues mutated in the BA.1 RBD, S371L, S373P and S375F in the loop adjacent to the lipid binding pocket, are also mutated in BA.2, but the mutation at 371 is Phe, meaning that this is likely to be a single point mutation for the early pandemic, whereas the S317L mutation in BA.1 requires two mutations. Thus, BA.2 can have features in common with earlier versions of the Omicron lineage. In addition, the various views of this part of the structure provided show that it adopts a range of different conformations. This is likely to be due to different crystal contacts and reflects the flexibility of this loop region. This can have biological function as the Ser mutations require a double codon change and can affect RBD presentation. As we have multiple views of this loop in the early pandemic virus, VoCs, Omicron BA.1 and BA.2, we can see that the flexibility is maintained in all variants.

[0427] Modeling the impact of BA.1, BA.1.1 and BA.2 changes on selected commercial known antibodies, early pandemic and Beta mAbs

[0428] Antibodies A (REGN 10987) and B (10933) are known: Antibody B (REGN 10933) binds to the posterior part of the left shoulder, and REGN 10987 binds to the right shoulder. Both are knocked out by the Omicron lineage, except for known antibodies A (REGN 10987) and BA.2. Antibody B (REGN 10933) H2 contacts residue 493, and due to the presence of Q493R in all Omicron strains, neutralizing activity against Omicron is generally lost. Antibody A (REGN 10987) H2 contacts residue 446. BA.2 uniquely lacks the G446S mutation, therefore regn10987 retains some neutralizing ability.

[0429] Antibodies C (AZD1061) and D (AZD8895) are known to bind to the posterior region of the left shoulder and the anterior region of the right shoulder, respectively, both showing reduced neutralization. Antibody C (AZD1061) is still able to neutralize BA.2 and BA.3 (reduced by approximately 10-fold), but compared to Victoria, the neutralization of BA.1 is reduced by >100-fold, and the neutralization of BA.1.1 is reduced by >1000-fold compared to Victoria. Antibody C (AZD1061) is known to be affected by contact with the G446S mutation (absent in BA.2 and BA.3) and the R346K (BA.1.1) mutation (contacted by L2 and H3). Antibody D (AZD8895) is known to be a VH1-58 antibody, contacting residues 477(H3) and 493(H2), and is compromised by the S477N and Q493R mutations prevalent in the Omicron lineage. Antibody E (AZD7442) (a combination of C and D) is known to maintain some neutralizing activity against Omicron strains as a whole.

[0430] Known antibodies F, G, and H: All known antibodies F, G, and H exhibit a significant loss of activity against Omicron. Known antibodies F and H show complete loss of activity, while known antibody G (ADG20) shows a 276-fold reduction in activity against Omicron.

[0431] Known antibodies I and J: Both antibodies have their activity knocked out across the entire Omicron lineage. Known antibody J (Ly-CoV16) (VH3-53) makes extensive interactions with N501 and Y505 via LI and L3, making it sensitive to mutations at these residues. Known antibody I (Ly-CoV-555) is susceptible to the E484K mutation in delta, but can be resistant to E484A, but it also contacts residue 493, so the ubiquitous Omicron Q493R mutation will completely abolish binding.

[0432] Known antibody K (S309): Known antibody K (S309) retains reasonable activity across the entire Omicron lineage. S309 contacts H3 on the right wing which contacts G339 and N343 glycans, which are close to the serine 371, 373 and 375 mutations. The S371F mutation in BA.2 (but not S371L) can impact binding, leading to slightly diminished activity against this virus.

[0433] Structure of BA.2 RBD and ACE2 affinity

[0434] Omicron BA.1, BA.1.1, BA.2 and BA.3 RBD affinity for ACE2 was measured by SPR and BLI. The affinity of BA.1 was comparable to that of the earlier viruses, 8nM and 7nM respectively (binding affinities of Omicron RBDs are shown in Tables 14 and 18), meaning that the increased affinity conferred by S477N, Q498R and N501Y was offset by other mutations in the ACE2 footprint. The affinity of BA.2 was slightly increased compared to the earlier viruses (about 1.5x and Y nM respectively). Based on earlier measurements of the contribution of individual mutations to binding affinity, G496S, as well as the triple mutations S371L, S373P and S375F reduced binding by 2-fold and 2.2-fold respectively, while BA.2 lacks G496S and has S371F. This can be part of the difference, but more likely is that mutations on the periphery of the ACE2 footprint in BA.2 can enhance binding. This is borne out by the structure of BA.2 / ACE2.

[0435] BA.4 / 5 RBD and ACE2 affinity

[0436] BA.4 / 5 RBD affinity for ACE2 was also measured by SPR Figure 12A-D). The affinity of the BA.4 / 5 RBD was increased compared to the ancestral virus, BA.1 and BA.2 (approximately 3-fold, 3-fold and 2-fold respectively (BA.4 / 5 / ACE2 KD = 2.4 nM)) (Deinirattisai et al., 2022; Nutalai et al., 2022), which was mainly due to an increase in the binding half-life. Modelling of the ACE2 / RBD complex indicated that most of this effect came from electrostatic complementarity between ACE2 and the RBD caused by the L452R mutation Figure 12 E-G).

[0437] BA.2.45 RBD and ACE2 affinity

[0438] Surface plasmon resonance (SPR) was also used to characterise the interaction between ACE2 and the BA.2.75 RBD. The off-rate was very slow, resulting in sub-nanomolar affinity (BA.2.75 / ACE2 KD = 0.45 nM) Figure 18 A, B). This represents a significant increase in affinity compared to BA.2 (9-fold) Figure 18 C), and even tighter than BA.4 / 5 (5-fold) Figure 18 D), which binds to ACE2 with higher affinity than BA.2 (Tuekprakhon et al., 2022). BA.2.75 was found to be the strongest ACE2 binder among all SARS-CoV-2 VoCs, including Alpha (Alpha / ACE2 KD = 1.5 nM; Figure 18 E), and the first SARS-CoV-2 VoC with sub-nanomolar affinity.

[0439] The BA.2 + N460K RBD could not be expressed, but the binding affinity of the BA.2 + R493Q RBD to ACE2 was also measured Figure 18 F) (KD = 0.55 nM). This confirmed that the R493Q back-mutation contributes to the high affinity of the BA.2.75 RBD.

[0440] The impact of mutations in BA.2.75

[0441] The group of mutations in BA.2.75 had opposite effects on neutralization compared to BA.2. The back mutation R493Q makes the virus more susceptible to neutralization by vaccine sera (vaccines contain Q493), while N460K, in isolation, reduces neutralization titers to a greater extent than the combination of mutations observed in BA.2.75. N460K is a novel substitution that has not appeared in previous SARS-CoV-2 variants. This mutation was introduced into the BA.2 backbone and its effect on neutralization by BNT162b2 sera was evaluated. Strikingly, BA.2+N460K titers were reduced by 3.1-fold compared to BA.2, greater than the reduction observed in BA.2.75 and comparable to that observed in BA.4 / 5.

[0442] Using a panel of potent mAbs derived from vaccinees who experienced a BA.1 vaccine breakthrough infection, it was shown that many mAbs belonging to the IGHV3-53 / 66 family had reduced or were knocked out for activity against BA.2.75. IGHV3-53 / 66 is the most commonly isolated mAb in SARS-CoV-2 and binds to an epitope on the “neck”. Thus, IGHV53 / 66 forms a major public antibody response and it is not surprising that the virus has evolved to evade this response.

[0443] Although the BA.2+N460K RBD cannot be expressed, a previous study using yeast display showed that N460K can enhance RBD binding to ACE2, an effect similar to the N501Y mutation first described in Alpha (Zahradnik et al., 2021). Thus, N460K can enhance antibody escape as well as increase receptor binding affinity.

[0444] Interestingly, BA.2.75 also acquired the R493Q back mutation (Q493R was acquired in BA.1 and is present in all other Omicron sublineages except BA.4 / 5). The BA.2.75 RBD is able to bind to ACE2 with 9-fold higher affinity than BA.2 and more tightly than BA.4 / 5 (Dejnirattisai et al., 2022; Tuekprakhon et al., 2022). This is partly due to the R493Q mutation. The BA.2.75 RBD has the highest receptor binding affinity of all SARS-CoV-2 variants measured to date.

[0445] These data suggest that there can be a delicate balance between antibody escape and ACE2 receptor affinity. Mutations in BA.2.75 lead to reduced neutralization titers of vaccine sera compared to BA.2. Single BA.2.75 mutations can lead to more reduction in neutralization titers compared to the full BA.2.75 S sequence, but these mutations are balanced by the R393Q back mutation that has been selected to increase affinity to ACE2 and increase the transmissibility of BA.2.75.

[0446] BA.2.11, BA.2.12 and BA.2.13 RBD and ACE2 affinity

[0447] To assess the possible changes in transmissibility of BA.2 subvariants, SPR experiments were performed to analyze their RBD binding to ACE2 Figure 27 d-g). The affinity of these three RBD variants to ACE2 is approximately 3 nM, slightly higher than previously reported for BA.2 RBD (KD = 4 nM) (Nutalai et al., 2022). Modeling of the ACE2 / RBD complex indicates that this increase in affinity can be due to the mutation at leucine 452 that slightly improves the complementarity between ACE2 and RBD. Thus, these variants can have a subtle advantage over BA.2 in terms of transmission.

[0448] Antigenic mapping of BA.3 and BA.4 / 5

[0449] BA.3 and BA.4 / 5 have been placed on an antigenic map using the neutralization data described above. The approach used to analyze Delta and Omicron variants (Liu et al., 2021, “Reduced neutralization of SARS-CoV-2 B.1.617 by vaccine and convalescent serum”. Cell 184, 4220-4236 e4213) was repeated, where independent modeling of individual viruses allowed for serum-specific scaling of the reactions. The reactions measured and modeled are shown in Figure 13 A (with 1551 observations and 340 parameters, with a residual error of 23%). The results are best visualized in three dimensions (see Figure 13 2D projections of B). This shows that, as expected, the Omicron sublineage clusters together, but is completely separate from earlier pandemic viruses and the earlier VoCs. Within the Omicron cluster, BA.4 / 5 is the furthest from the earlier Omicron viruses.

[0450] Antigenic mapping of BA.2.75

[0451] BA.2.75 was tested for neutralization using sera from individuals previously infected during the pandemic. These sera included sera obtained early in the pandemic (prior to the emergence of Alpha), as well as sera obtained following infection with Alpha, Beta, Gamma, Delta, BA.1, and BA.2 Figure 25 As expected, BA.2.75 neutralization titers were lower than homologous infection strains (e.g., Alpha sera for Alpha virus). However, most striking was the complete loss of BA.2.75 neutralization using Delta sera (zero samples reached 50% neutralization at 1 / 20 dilution). However, titers against BA.2.75 were much higher in cases that were vaccinated prior to or after Delta infection.

[0452] Using these data, BA.2.75 was placed on a three-dimensional antigen map using the method previously reported in Tuekprakhon et al., 2022 Figure 22 A, B). Initially, all VoCs were included Figure 22 A); this indicated that BA.2.75 was grouped with other Omicron viruses, which were isolated into one hemisphere of the 3D map. BA.2.75 appeared to be completely separate from other Omicron sublineages, particularly BA.4 / 5. Also notable was that BA.2.75 and Delta were directly opposite in the map, emphasizing the antigenic distance between these two viruses. Since the data are high-dimensional, the 3D projection can distort true distances, so calculations were performed only for Omicron and early pandemic viruses (but preserving the full serological information for each virus). Results are shown in Figure 22 B and summarized the main features of the full map, but allowed for more extensive distribution of the Omicron sublineages in 3D space. It is notable that if the aggregated early pandemic and BA.2 / BA.3 pairs were combined, then these points distributed as a trigonal bipyramid, which maximized their separation, consistent with antigenic escape being an important factor in their evolution.

[0453] Example 3. Examples of antibodies that can be generated by swapping the light chain between antibodies derived from the same heavy chain V gene

[0454] As discussed in detail above, antibodies derived from the same heavy chain V gene can be swapped for light chains to produce antibodies comprising the heavy chain variable region of a first antibody and the light chain variable region of a second antibody, and such new antibodies can have improved neutralization and / or other properties when compared to the "parent" antibodies.

[0455] Tables 4 through 12 provide examples of such antibodies that can be generated by swapping the light chains between antibodies derived from the same heavy chain V gene. Table 17 provides information on the heavy and light chain V genes from which 28 Omicron-specific mAbs originate, and their specificity for the RBD or NTD of the SARS-CoV-2 spike protein.

[0456] Example 4. Materials and methods

[0457] Virus stock solution

[0458] SARS-CoV-2 / human / AUS / VIC01 / 2020 (Caly et al., 2020) Alpha and Beta were provided by Public Health England, Gamma was cultured from throat swabs from Brazil, Delta was donated by Wendy Barclay and Thushan de Silva from the UK G2P Genotyping to Phenotyping Consortium, and Omicron was cultured from positive throat swabs (IRAS Project ID: 269573, Ethics Reference No.: 19 / NW / 0730). In summary, VeroE6 / TMPRSS2 cells (NIBSCs) were maintained at 37°C in the presence of 5% CO2 in DMEM high-glucose supplemented with 1% fetal bovine serum, 2 mM Glutamax, 100 IU / ml penicillin-streptomycin, and 2.5 μg / ml amphotericin B, followed by 200 μL liquid inoculation with swabs. Cells were further maintained at 37°C, and cytopathic effect (CPE) was observed daily. Virus-containing supernatants were clarified at 80% CPE by centrifugation at 3,000 rpm at 4°C, and then stored as single-use aliquots at -80°C. Viral titers were determined by lesion formation assays on Vero CCL-81 cells (ATCC).

[0459] Sequencing of the Omicron isolates revealed the expected shared S gene variations (A67V, Δ69-70, T95I, G142D / Δ143-145, Δ211 / L212I, ins214EPE, G339D, S371L, S373P, S375F, K417N, N440K, G446S, S477N, T478K, E484A, Q493R, G496S, Q498R, N501Y, Y505H, T547K, D614G, H655Y, N679K, P681H, N764K, D796Y, N856K, Q954H, N969K, L981F), a complete furin cleavage site, and a single additional mutation, A701V.

[0460] Cells were infected with SARS-CoV-2 virus using a MOI of 0.0001.

[0461] Virus containing supernatant was harvested at 80% CPE and spun at 3000 rpm at 4°C and then stored at -80°C. Virus titers were determined by performing focus forming assays on Vero cells. Victoria passage 5, Alpha passage 2, and Beta passage 4 stocks Gamma passage 1, Delta passage 3, and Omicron passage 1 were sequenced to verify they contained the expected spike protein sequence and that the furin cleavage site had not changed.

[0462] Bacterial strains and cell culture

[0463] Vero (ATCC CCL-81) and VeroE6 / TMPRSS2 cells were cultured at 37°C in Dulbecco’s Modified Eagle Medium (DMEM) high glucose (Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS), 2mM GlutaMAX (Gibco, 35050061), and 100 U / ml penicillin-streptomycin. Human mAbs were expressed in HEK293T cells cultured at 37°C and 5% CO2 in UltraDOMA PF protein free media (cat# 12-727F, LONZA). HEK293T (ATCC CRL-11268) cells were cultured at 37°C and 5% CO2 in DMEM high glucose (Sigma-Aldrich) supplemented with 10% FBS, 1% 100X Mem Neaa (Gibco), and 1% 100X L-glutamine (Gibco). For expression of RBD, RBD variants, and ACE2, HEK293T cells were cultured at 37°C in DMEM high glucose (Sigma) supplemented with 2% FBS, 1% 100X Mem Neaa, and 1% 100X L-glutamine for transfection. Omicron RBD and human mAbs were also expressed in HEK293T (ATCC CRL-11268) cells cultured at 37°C and 5% CO2 in FreeStyle 293 expression medium (ThermoFisher, 12338018). E. coli DH5a bacteria were used for transformation and large scale preparation of plasmids. Single colonies were picked and cultured in LB broth at 200 rpm in a shaker at 37°C overnight.

[0464] Serum from Pfizer vaccinees

[0465] Pfizer vaccine sera were obtained from volunteers who had received one or two doses of the BNT162b2 vaccine. The vaccinees were healthcare staff at the Oxford University Hospitals NHS Foundation Trust, had not previously been infected with SARS-CoV-2, and participated in the OPTIC study as part of the Oxford Biomedical Research Unit for Gastrointestinal and Liver (Biomedical Research Unit for Gastrointestinal and Liver) Library Study 16 / YH / 0247 [Yorkshire & The Humber - Sheffield Research Ethics Committee (REC)] which had been amended for this on 8th June 2020. The study was conducted in accordance with the principles of the Declaration of Helsinki (2008) and International Conference on Harmonization (ICH) Good Clinical Practice (GCP) guidelines. All participants who took part in the study gave written informed consent. Participants were studied after receiving two doses of the Pfizer / BioNTech BNT162b2 mRNA vaccine (30 microgram, diluted intramuscular injection (0.3 mL each)) and sampled approximately 28 days (range 25-38 days), 180 days (range 178-221 days) and 270 days (range 243-273 days) after receiving two doses of the vaccine, then 17-28 days after receiving a third booster dose of BNT162B2 vaccine, then approximately 28 days (range 25-56 days) after. The mean age of the vaccinees was 37 years (range 22-66 years), 21 males and 35 females.

[0466] Plasma from early pandemic and Alpha cases

[0467] Participants from the first wave of SARS-CoV2 in the UK and participants with sequence-confirmed B.1.1.7 lineage in December 2020 and February 2021 were recruited through three studies: Sepsis Immunomics [Oxford REC C, reference: 19 / SC / 0296], ISARIC / WHO Clinical Characteristics of Severe Illness in COVID-19 Protocol [Oxford REC C, reference: 13 / SC / 0149] and Oxford Gastroenterology: COVID Substudy [Sheffield REC, reference: 16 / YH / 0247]. Diagnosis was confirmed by reporting symptoms consistent with COVID-19 and detection of SARS-CoV-2 positive from upper respiratory (nasal / throat) swabs tested using reverse transcriptase polymerase chain reaction (RT-PCR) from accredited laboratories. Blood samples were taken at least 14 days after symptom onset with consent. Clinical information was captured for all individuals at the time of sampling, including severity of illness (classified as mild, severe or critical infection according to WHO recommendations) and time between symptom onset and sampling and age of the participant. After heat inactivation of plasma / serum samples, they were aliquoted to allow no more than 3 freeze-thaw cycles for data generation.

[0468] Serum from Beta, Gamma and Delta and BA.1 infection cases

[0469] Beta and Delta samples from UK infection cases were collected under the “Innate and adaptive immunity to SARS-CoV-2 in healthcare workers and household members” protocol, which is attached to the Oxford Gastroenterology: COVID Substudy, discussed above and approved by the Central University Research Ethics Committee, University of Oxford. All individuals had sequence-confirmed Beta / Delta infection or PCR-confirmed symptomatic illness, which occurred during isolation and direct contact with Beta / Delta sequence-confirmed cases. Additional Beta infection serum (sequence-confirmed) was obtained from South Africa. At the time of swab collection, patients signed informed consent to agree to the collection of data and serial blood samples. The study was approved by the University of the Witwatersrand Human Research Ethics Committee (reference number 200313) and conducted in accordance with the Good Clinical Practice guidelines. Gamma samples were provided by the Coronavirus International Reference Laboratory at FIOCRUZ (WHO) as part of the national coronavirus surveillance and were approved by the FIOCRUZ Ethics Committee (CEP 4.128.241) to receive and analyze samples from suspected COVID-19 cases for virological surveillance. Clinical samples were shared with the University of Oxford, UK, under MTA IOC FIOCRUZ 21-02.

[0470] Serum from BA.1 infection cases, study subjects

[0471] Individuals carrying omicron BA.1 were enrolled in ISARIC / WHO Clinical Characterisation Protocol for Severe Novel Infection [Oxford REC C, reference: 13 / SC / 0149] and the “Healthcare worker household and household member innate and adaptive immunity to SARS-CoV-2” protocol attached to the Oxford Gastrointestinal Disease: COVID sub-study [Sheffield REC, reference: 16 / YH / 0247] further approved by the University of Oxford Central University Research Ethics Committee, following informed consent. Diagnosis was confirmed by reporting symptoms consistent with COVID-19 or positive contact with a known Omicron case and detection of SARS-CoV-2 positive and lineage sequence confirmed by the National Reference Laboratory from upper respiratory (nasal / throat) swabs tested using reverse transcriptase polymerase chain reaction (RT-PCR) from accredited laboratories. Blood samples were taken at least 10 days after PCR test confirmation, following consent. Clinical information was captured at the time of sampling for all individuals, including severity of illness (classified as mild, severe or critical infection according to WHO recommendations) and time between symptom onset and sampling and age of the participant.

[0472] AstraZeneca-Oxford vaccine study procedures and sample handling

[0473] Full details of the randomised controlled trials of ChAdOxl nCoV-19 (AZD1222) have been published previously (PMID: 33220855 / PMID: 32702298). These studies were registered with ISRCTN (15281137 and 89951424) and ClinicalTrials.gov (NCT04324606 and NCT04400838). Copies of the protocols are included in previous publications (Folegatti et al., 2020, Lancet 396, 467-478).

[0474] Data from vaccinated volunteers who received two vaccinations are included in the examples. Vaccine doses were 5 x 10 10 viral particles (standard dose; SD / SD cohort n=21) or half dose as their first dose (low dose) and standard dose as their second dose (LD / SD cohort n=4). The interval between first and second dose was in the range 8-14 weeks. Blood samples were collected and serum isolated on the day of vaccination and pre-specified days after vaccination (e.g. 14 and 28 days post-boost).

[0475] Focus reduction neutralisation test (FRNT)

[0476] The neutralizing potential of Abs was measured using a focus reduction neutralization test (FRNT), in which the reduction in the number of infection foci was compared to negative control wells without antibody. Briefly, serially diluted Abs or plasma were mixed with SARS-CoV-2 strain and incubated for 1 hour at 37°C. The mixture was then transferred in duplicate to 96-well cell culture treated flat bottom microplates containing confluent Vero cell monolayers and incubated for an additional 2 hours, after which 1.5% semi-solid carboxymethylcellulose (CMC) overlay medium was added to each well to restrict virus spread. The focus-forming assay was then performed by staining the Vero cells with human anti-NP mAb (mAb206) followed by peroxidase-conjugated goat anti-human IgG (A0170; Sigma) in sequence. Finally, approximately 100 foci (infected cells) per well in the absence of antibody were visualized by the addition of TrueBlue peroxidase substrate. Virus-infected cell foci were counted on a classic AID EliSpot reader using AID ELISpot software. The probit program from the SPSS software package was used to calculate the percentage of focus reduction and determine the IC50.

[0477] Plasmid construction and pseudotyped lentiviral particle production

[0478] Pseudotyped lentiviruses expressing SARS-CoV-2 S protein from ancestral strain (Victoria, S247R), BA.1, BA.1.1, and BA.2 were constructed as previously described (Nie, Jianhui, et al. “Establishment and validation of a pseudovirus neutralization assay for SARS-CoV-2.” Emerging microbes & infections 9.1 (2020): 680-686; Liu, Chang, et al. “Reduced neutralization of SARS-CoV-2 B.1.617 by vaccine and convalescent serum.” Cell 184.16 (2021): 4220-4236) with some modifications. Briefly, synthetic codon-optimized SARS-CoV-2 BA.1 and BA.2 were custom synthesized by GeneArt (Thermo Fisher Scientific GENEART). Insert and pcDNA3.1 vectors were cloned by using Gibson assembly. Victoria (S247R) construct was as previously described in Liu, Chang, et al. “Reduced neutralization of SARS-CoV-2 B.1.617 by vaccine and convalescent serum.” Cell 184.16 (2021): 4220-4236.To construct BA.1.1, mutagenic primers for R346K (R346K_F 5'-GTGTTCAATGCCACCAAATTCGCCAGCGTGTAC-3' and R346K_R 5'-GTACACGCTGGCGAATTTGGTGGCATTGAACAC-3') were PCR amplified by using the BA.1 construct as a template along with two primers for the pcDNA3.1 vector (pcDNA3.1 BamHI_F 5'-GGATCCATGTTCCTGCTGACCACCAAG AG-3' and pcDNA3.1_Tag_S_EcoRI_R 5'-GAATTCTCACTTCTCGAACTGAGGGTGGC-3'), purified using QIAquick Gel Extraction Kit (QIAGEN), and then Gibson assembled. After isolating the plasmids using QIAGEN Miniprep Kit (QIAGEN), all constructs were verified by Sanger sequencing.

[0479] A similar strategy was applied for BA.3 and BA.4 / 5, briefly, BA.3 mutations were constructed using combined fragments from BA.1 and BA.2. The resulting mutations were as follows. The resulting mutations were as follows: A67V, Δ69-70, T95I, G142D, Δ143-145, Δ211 / L212I, G339D, S371F, S373P, S375F, D405N, K417N, N440K, G446S, S477N, T478K, E484A, Q493R, Q498R, N501Y, Y505H, D614G, H655Y, N679K, P681H, N764K, D796Y, Q954H, and N969K. Although BA.4 / 5 S protein shares some amino acid mutations with BA.2 (Nutalai et al., 2022), to generate BA.4 / 5 mutations, Δ69-70, i.e., L452R, F486V, and R498Q were added. The resulting pcDNA3.1 carrying the S gene was used to generate pseudovirus particles along with lentiviral packaging vectors and a transfer vector encoding luciferase reporter. Sequence confirmation was performed for the integrity of the constructs.

[0480] BA.2.12.1 was constructed using the same method, and BA.2.75 was constructed by adding more mutations to BA.2. To generate BA.2.75, K147E, W152R, F157L, I210V, G275S, G446S, and N460K were added to the BA.2 backbone. 339D was also changed to 339H in BA.2 S, and 493R was reverted to 493Q in BA.2, identical to the ancestral strain. To test the effect of individual mutations, D339H, G446S, N460K, and R493Q were introduced individually into the BA.2 backbone. The resulting pcDNA3.1 plasmid carrying the S gene, along with a lentiviral packaging vector and a transfer vector encoding a luciferase reporter, was used to generate pseudovirus particles.

[0481] fake virus neutralization test

[0482] The details of the pseudovirus neutralization test were previously described (Liu, Chang et al., “Reduced neutralization of SARS-CoV-2 B.1.617 by vaccine and convalescent serum.” Cell 184.16(2021): 4220-4236), with some modifications. In short, the neutralizing activity of potent monoclonal antibodies (mAbs) produced from donors who had recovered from Omicron and Beta infections, as well as those infected during the early stages of the UK pandemic, against Victoria, Omicron-BA.1, BA.1.1, BA.2, BA.2.11, BA.2.12.1, BA.2.13, BA.3, BA.4.6, BA.4 / 5, BA.2.75, and BA.2+N460K was determined. Each mAb was serially diluted four times and incubated with pseudovirus particles at 37°C and 5% CO2 for 1 hour. Stable HEK293T / 17 cells expressing human ACE2 were then added to the mixture at 1.5 x 10⁴ cells / well. Forty-eight hours post-transduction, the culture supernatant was removed, and 50 μL of 1x PBS solution from a 1:2 Bright-Glo™ luciferase assay system (Promega, USA) was added to each well. The reaction was incubated at room temperature for 5 minutes and then... Firefly luciferase activity was measured (BMG Labtech, Ortenberg, Germany). The percentage of mAb neutralization was calculated relative to the control. Probit analysis was used to estimate the dilution value (PVNT50) that inhibited half of the maximally pseudotyped lentivirus infection.

[0483] To determine the neutralizing activity of the convalescent plasma / serum samples or vaccine sera, 3-fold serial dilutions of the samples were incubated with pseudovirus particles for 1 hour and the same strategy as for the mAbs was adopted.

[0484] DNA manipulation

[0485] Cloning was performed by using the unrestricted method (Peleg and Unger, 2014). The mutagenic megaprimer was PCR amplified (KAPA HiFi HotStart ReadyMix, Roche, Switzerland, Cat# KK3605) by using purified with the gel and PCR purification kit (Nacherey-Nagel, Germany, REF 740609.50) and cloned into pJYDC1 (Adgene ID: 162458) (Zahradnik et al., 2021a). The parental pJYDC1 molecule was cut by Dpnl treatment (1 h, NEB, USA, Cat# R0176) and the reaction mixture was electroporated into E. coli 10G cells (Lucigen, USA). The correctness of the mutagenesis was verified by sequencing.

[0486] Cloning of Spike protein and RBD

[0487] Expression plasmids for wild type and Omicron BA.1 Spike protein and RBD of BA.1 and BA.2 encoding the human codon-optimized sequences of BA.1 (EPI_ISL_6640917) and BA.2 (EPI_ISL_6795834.2) were constructed. The construct of the wild type and BA.1 Spike protein and RBD plasmids was identical to previously described (Dejnirattisai, Wanwisa et al. “The antigenic anatomy of SARS-CoV-2 receptor binding domain.” Cell 184.8 (2021): 2183-2200). The synthetic codon-optimized RBD fragment of BA.2 was used as a template and the construct was amplified by PCR and cloned into the pNEO vector as previously described (Dejnirattisai et al., 2021a; Supasa et al., 2021; Zhou et al., 2021). The construct was verified by Sanger sequencing.

[0488] To generate His-tagged BA.4 / 5 RBD constructs, site-directed PCR mutagenesis was performed using the BA.2 RBD construct as a template (Nutalai et al., 2022) with the introduction of L452R, F486V, and R493Q mutations. The gene fragments were amplified with pNeoRBD333Qmi|F (5’-GGTTGCGTAGC TGAAACCGGTCATCACCATCACCATCACACCAATCTGTGCCCT TTCGAC-3’) and pNeoRBD333_R (5’-GTGATGGTGGTGCTTGGTACCT TATTACTTCT TGCCGCACACGGTAGC-3’) and cloned into the pNeo vector (Supasa et al., 2021, “Reduced neutralization of SARS-CoV-2 B.1.1.7 variant by convalescent and vaccine sera”. Cell 184, 2201-2211 e2207). To generate BA.4 / 5 RBD constructs containing the BAP-His tag, the gene fragments were amplified with RBD333F (5’-GCGTAGCTGAAACCGGCACCAATCT GTGCCCTTTCGAC-3’) and RBD333BAP R (5’-GTCATTCAGCAAGC TCTTCTTGCCGCACACGG TAGC-3’) and cloned into the pOPINTTGneo-BAP vector (Huo et al., 2020, “Neutralizing nanobodies bind SARS-CoV-2 spike RBD and block interaction with ACE2”. Nature structural & molecular biology 27, 846-854). Cloning was performed using the ClonExpress II One-Step Cloning Kit (Vazyme). After isolating plasmids using the QIAGEN Miniprep Kit (QIAGEN), the constructs were verified by Sanger sequencing.

[0489] To generate BA.2.75 RBD constructs, site-directed PCR mutagenesis was performed using the BA.2 spike protein construct as a template (Nutalai et al., 2022) with the introduction of Figure 26The primers listed in the text introduce D339H, G446S, N460K, and R493Q mutations; using D339H_pNeoF and RBD333_BAP_R ( Figure 26 The gene fragment was amplified and cloned into the pOPINTTGneo-BAP vector (Huo et al., 2020, "Neutralizing nanobodies bind SARS-CoV-2 spike RBD and block interaction with ACE2." Nature structural & molecular biology 27, 846-854). To generate the BA.2+R493Q RBD construct, the BA.2 spike protein construct was used as a template for site-directed PCR mutagenesis, in which... Figure 26 The primers listed were used to introduce the R493Q mutation; the gene fragment was amplified using pNeoRBD333Omi_F and BD333_BAP_R and cloned into the pNeo vector (Supasa et al., 2021, “Reduced neutralization of SARS-CoV-2 B.1.1.7 variant by convalescent and vaccine sera.” Cell 184, 2201-2211e2207). Cloning was performed using the ClonExpress II one-step cloning kit (Vazyme). After plasmid isolation using the QIAGEN Miniprep kit (Kiagene), the construct was validated by Sanger sequencing.

[0490] The generation of RBD

[0491] The plasmid encoding RBD was transfected into Expi293FT via PEI. M In cells (Thermo Fisher Scientific), in FreeStyle TM The cells were cultured for 4 days at 30°C and 8% CO2 in 293 expression medium (Thermo Fisher Scientific). To express biotinylated RBD, the RBD-BAP plasmid was co-transfected with pDisplay-BirA-ER (Addgene plasmid 20856; encoding a biotinylate ligase for ER localization) in the presence of 0.8 mM D-biotin (Sigma-Aldrich).

[0492] The generation of BA.2.75 RBD

[0493] The plasmid encoding RBD was transfected into Expi293FT via PEI. M In cells (Thermo Fisher Scientific), in FreeStyle TM The cells were cultured in 293 expression medium (Thermo Fisher Scientific) at 37°C and 8% CO2 for 1 day, followed by 30°C for 3 days. To express biotinylated RBD, the RBD-BAP plasmid was co-transfected with pDisplay-BirA-ER (Addgene plasmid 20856; encoding a biotinylate ligase for ER localization) in the presence of 0.8 mM D-biotin (Sigma-Aldrich). Conditioning medium was diluted 1:2 to binding buffer (50 mM sodium phosphate, 500 mM sodium chloride, pH 8.0). RBD was purified via His-tag binding using a 5 mL HisTrap nickel column (GE Healthcare), followed by purification using a Superdex 7510 / 300GL gel filter column (GE Healthcare) in 10 mM HEPES and 150 mM sodium chloride.

[0494] protein production

[0495] Protein expression and purification were performed as previously described (Dejnirattisai et al., 2021a; Zhou et al., 2020). Briefly, the plasmid encoding the protein was transiently expressed in HEK293T (ATCCCRL-11268) cells. Conditioned media were concentrated using a QuixStand benchtop system. His-labeled Omicron RBD was purified using a 5 mL HisTrap nickel column (GE Healthcare) and further purified using a Superdex 75HiLoad 16 / 60 gel filter column (GE Healthcare). Streptococcus-labeled Omicron spike protein was purified using Strep-Tactin XT resin (IBA Lifesciences). Approximately 4 mg of ACE2 was mixed with a self-made His-labeled 3C protease and DTT (final concentration 1 mM). After incubation at 4°C for one day, the sample was passed through a 5 mL HisTrap nickel column (GE Healthcare). The His-labeled protein was removed by passing it through the nickel column, and the purified ACE2 was harvested and concentrated.

[0496] IgG mAb and Fab purification

[0497] For purification of full-length IgG mAbs, supernatant from mAb expression was collected and filtered through a vacuum filtration system and loaded onto protein A / G beads at 4°C overnight. The beads were washed three times with PBS and the IgG was eluted using 0.1 M glycine pH 2.7. The eluate was neutralized with Tris-HCl pH 8 buffer to a final pH = 7. The IgG concentration was determined by spectrophotometry and buffer exchanged into PBS. For expression and purification of Fab 158 and EY6A, the heavy and light chain expression plasmids for the Fab were co-transfected into HEK293T cells by PEI. After culturing the cells for 5 days at 37°C and 5% CO2, the culture supernatant was harvested and filtered using a 0.22 mm polyether sulfone filter. Fab 158 was purified using Strep-Tactin XT resin (IBA Lifesciences) and Fab EY6A was purified with a Ni-NTA column (GE Healthcare) and a Superdex 75 HiLoad 16 / 60 gel filtration column (GE Healthcare). AstraZeneca and Regeneron antibodies were provided by AstraZeneca and Vir, Lilly and Adagio antibodies were provided by Adagio. For these antibodies, the heavy and light chains of the indicated antibodies were transiently transfected into 293Y cells and the antibodies were purified from the supernatant on protein A. The Fab fragments of 58 and β-55 were digested from purified IgG with papain using the Pierce Fab Preparation Kit (Thermo Fisher) following the manufacturer’s protocol.

[0498] Surface plasmon resonance

[0499] Surface plasmon resonance experiments were performed using a Biacore T200 (GE Healthcare). All assays were performed at 25°C with a running buffer of HBS-EP (GE Healthcare).

[0500] To determine the binding kinetics between SARS-CoV-2 RBD and ACE2 / monoclonal antibodies (mAbs), a protein A sensor chip (GE Healthcare) was used. ACE2-Fc or mAbs were immobilized onto the sample flow cell of the sensor chip. The reference flow cell was left empty. Using a single cycle kinetics program, a series of five concentrations of RBD prepared by serial two-fold dilutions were injected into both flow cells at a flow rate of 30 μΐ min -1 All data were fitted to a 1 : 1 binding model using Biacore T200 Evaluation Software 3.1.

[0501] To determine the binding kinetics between SARS-CoV-2 Spike protein and ACE2, a CM5 sensor chip was used. The sensor chip was first activated by injecting an equal volume of EDC and NHS (GE Healthcare) mixture at 20 uL / min for 300 s, then 20 ug / mL of Spike protein sample (GE Healthcare) in 10 mM sodium acetate pH 5.0 was injected at 10 uL / min onto the sample flow cell of the sensor chip, and finally 1.0 M ethanolamine-HCl pH 8.5 (GE Healthcare) was injected at 20 uL / min for 180 s. The reference flow cell was left blank. A single cycle kinetics program was used to inject ACE2 at a flow rate of 30 μl min-1in a series of five concentrations prepared by serial twofold dilutions into both flow cells. Running buffer was also injected using the same program to subtract background. -1 To determine the binding kinetics between SARS-CoV-2 Spike protein and ACE2, a CM5 sensor chip was used. The sensor chip was first activated by injecting an equal volume of EDC and NHS (GE Healthcare) mixture at 20 uL / min for 300 s, then 20 ug / mL of Spike protein sample (GE Healthcare) in 10 mM sodium acetate pH 5.0 was injected at 10 uL / min onto the sample flow cell of the sensor chip, and finally 1.0 M ethanolamine-HCl pH 8.5 (GE Healthcare) was injected at 20 uL / min for 180 s. The reference flow cell was left blank. A single cycle kinetics program was used to inject ACE2 at a flow rate of 30 μl min-1in a series of five concentrations prepared by serial twofold dilutions into both flow cells. Running buffer was also injected using the same program to subtract background.

[0502] To determine the binding kinetics between SARS-CoV-2 Spike protein and ACE2, a CM5 sensor chip was used. The sensor chip was first activated by injecting an equal volume of EDC and NHS (GE Healthcare) mixture at 20 uL / min for 300 s, then 20 ug / mL of Spike protein sample (GE Healthcare) in 10 mM sodium acetate pH 5.0 was injected at 10 uL / min onto the sample flow cell of the sensor chip, and finally 1.0 M ethanolamine-HCl pH 8.5 (GE Healthcare) was injected at 20 uL / min for 180 s. The reference flow cell was left blank. A single cycle kinetics program was used to inject ACE2 at a flow rate of 30 μl min-1in a series of five concentrations prepared by serial twofold dilutions into both flow cells. Running buffer was also injected using the same program to subtract background.

[0503] To determine the binding kinetics between SARS-CoV-2 Spike protein and ACE2, a CM5 sensor chip was used. The sensor chip was first activated by injecting an equal volume of EDC and NHS (GE Healthcare) mixture at 20 uL / min for 300 s, then 20 ug / mL of Spike protein sample (GE Healthcare) in 10 mM sodium acetate pH 5.0 was injected at 10 uL / min onto the sample flow cell of the sensor chip, and finally 1.0 M ethanolamine-HCl pH 8.5 (GE Healthcare) was injected at 20 uL / min for 180 s. The reference flow cell was left blank. A single cycle kinetics program was used to inject ACE2 at a flow rate of 30 μl min-1in a series of five concentrations prepared by serial twofold dilutions into both flow cells. Running buffer was also injected using the same program to subtract background.

[0504] To compare the binding of BA.2 and BA.4 / 5 RBD to mAb Omi-06 / Omi-25 / Omi-26, a Protein A sensor chip (GE Healthcare) was used. The mAbs in IgG format were immobilized onto the sample flow cell of the sensor chip to similar levels (approximately 350 RU). The reference flow cell was left blank. A single RBD injection was performed at 200 nM on both flow cells at a flow rate of 30 pl min-1. Running buffer was also injected using the same program to subtract background. Sensorgrams were plotted using Prism 9 (GraphPad).

[0505] To compare the binding of BA.2 and BA.4 / 5 RBD to mAb Omi-02 / Omi-23 / Omi-31, a Biotin CAPture kit (GE Healthcare) was used. Biotinylated BA.2 and BA.4 / 5 RBD were immobilized onto the sample flow cell of the sensor chip to similar levels (approximately 120 RU). The reference flow cell was left blank. A single mAb Fab injection was performed at 200 nM on both flow cells at a flow rate of 30 pl min-1. Running buffer was also injected using the same program to subtract background. Sensorgrams were plotted using Prism 9 (GraphPad).

[0506] To determine the binding kinetics between BA.2.75 or BA.2+R493Q RBD and ACE2, a Protein A sensor chip (GE Healthcare) was used. ACE2-Fc was immobilized onto the sample flow cell of the sensor chip. The reference flow cell was left blank. A series of five concentrations prepared by serial two-fold dilutions of RBD were injected into both flow cells at a flow rate of 30 pl min-1using a single cycle kinetics program. Running buffer was also injected using the same program to subtract background. All data were fitted to a 1 : 1 binding model using Biacore T200 Evaluation Software 3.1.

[0507] To confirm the binding kinetics between BA.2.75 RBD and ACE2, a Biotin CAPture kit (GE Healthcare) was used. Biotinylated ACE2 (bio-ACE2) was immobilized onto the sample flow cell of the sensor chip. The reference flow cell was left blank. A series of five concentrations prepared by serial two-fold dilutions of BA.2.75 RBD were injected into both flow cells at a flow rate of 30 pl min -1 using a single cycle kinetics program. Running buffer was also injected using the same program to subtract background. All data were fitted to a 1 : 1 binding model using Biacore T200 Evaluation Software 3.1.

[0508] To determine the binding kinetics between BA.2.75 or BA.2 RBD and mAbs, a Biotin CAPture kit (GE Healthcare) was used. Biotinylated RBD was immobilized onto the sample flow cell of a sensor chip. The reference flow cell was left empty. Using a single-cycle kinetics program, a series of five concentrations of Omi-18 or Omi-32 Fab was injected into both flow cells at a flow rate of 30 pl min -1 To determine the binding kinetics between BA.2.75 or BA.2 RBD and mAbs, a Biotin CAPture kit (GE Healthcare) was used. Biotinylated RBD was immobilized onto the sample flow cell of a sensor chip. The reference flow cell was left empty. Using a single-cycle kinetics program, a series of five concentrations of Omi-18 or Omi-32 Fab was injected into both flow cells at a flow rate of 30 pl min -1 To determine the binding kinetics between BA.2.75 or BA.2 RBD and mAbs, a Biotin CAPture kit (GE Healthcare) was used. Biotinylated RBD was immobilized onto the sample flow cell of a sensor chip. The reference flow cell was left empty. Using a single-cycle kinetics program, a series of five concentrations of Omi-18 or Omi-32 Fab was injected into both flow cells at a flow rate of 30 pl min -1 To determine the binding kinetics between BA.2.75 or BA.2 RBD and mAbs, a Biotin CAPture kit (GE Healthcare) was used. Biotinylated RBD was immobilized onto the sample flow cell of a sensor chip. The reference flow cell was left empty. Using a single-cycle kinetics program, a series of five concentrations of Omi-18 or Omi-32 Fab was injected into both flow cells at a flow rate of 30 pl min

[0509] To compare the binding of BA.2 and BA.2.75 RBD to mAb Omi-29, a Biotin CAPture kit (GE Healthcare) was used. Biotinylated BA.2 and BA.2.75 RBD were immobilized onto the sample flow cell of a sensor chip, to similar levels (about 110 RU). The reference flow cell was left empty. A single mAb Fab injection was performed at 1 mM on both flow cells at a flow rate of 30 pl min -1 To compare the binding of BA.2 and BA.2.75 RBD to mAb Omi-29, a Biotin CAPture kit (GE Healthcare) was used. Biotinylated BA.2 and BA.2.75 RBD were immobilized onto the sample flow cell of a sensor chip, to similar levels (about 110 RU). The reference flow cell was left empty. A single mAb Fab injection was performed at 1 mM on both flow cells at a flow rate of 30 pl min

[0510] To compare the binding of BA.2 and BA.2.75 RBD to mAb Omi-29, a Biotin CAPture kit (GE Healthcare) was used. Biotinylated BA.2 and BA.2.75 RBD were immobilized onto the sample flow cell of a sensor chip, to similar levels (about 110 RU). The reference flow cell was left empty. A single mAb Fab injection was performed at 1 mM on both flow cells at a flow rate of 30 pl min -1 To compare the binding of BA.2 and BA.2.75 RBD to mAb Omi-29, a Biotin CAPture kit (GE Healthcare) was used. Biotinylated BA.2 and BA.2.75 RBD were immobilized onto the sample flow cell of a sensor chip, to similar levels (about 110 RU). The reference flow cell was left empty. A single mAb Fab injection was performed at 1 mM on both flow cells at a flow rate of 30 pl min

[0511] IgG mAb and Fab production

[0512] AstraZeneca and Regeneron antibodies were provided by AstraZeneca and Vir, Lilly and Adagio antibodies were provided by Adagio, LY-CoV1404 was provided by LifeArc. For in-house antibodies, the heavy and light chains of the indicated antibodies were transiently transfected into 293Y or 293T cells and antibodies were purified from supernatants on protein A as previously described (Nutalai et al., 2022). Fabs were digested from purified IgG with papain using the Pierce Fab Preparation Kit (Thermo Fisher Scientific) following the manufacturer’s protocol.

[0513] Quantification and statistical analysis

[0514] Statistical analysis is reported in the results and figure legends. Neutralisation was measured by FRNT. The probit program from the SPSS package was used to calculate the percentage of lesion reduction and to determine the IC50(FRNT50). Analysis was performed using the Wilcoxon matched-pairs signed-rank test and two-tailed P values were calculated from the geometric mean.

[0515] Crystallisation

[0516] RBD proteins were deglycosylated with endoglycosidase F1 prior to use for crystallisation. Omicron BA.1-RBD was mixed with Omi-12 and beta-54 Fabs at a 1 : 1 : 1 molar ratio, final concentration 7 mg ml-1, respectively. These complexes were incubated at room temperature for 30 minutes, respectively. Initial screening of crystals was set up using nanolitre sitting-drop vapour-diffusion method with a Cartesian robot in Crystalquick 96-well X-plates (Greiner Bio-One) with 100 nL protein plus 100 nL reservoir per drop, as previously described (Walter et al., 2003, Journal of Applied Crystallography 36, 308-314).

[0517] Crystals of BA.1-RBD / Omi-12 / beta-54 were formed under Hampton Research PEG Rx condition 1-46 containing 0.1 M trisodium citrate dihydrate pH 5.0 and 18% (w / v) PEG 20000. The complex of BA.1-RBD / Omi-12 / beta-54 was screened under Hampton Research Ammonium Sulfate Screen C2 containing 2.4 M (NH4)2SO4 and 0.1 M citric acid pH 5.0, but only crystals of the Fab Omi-12 alone were formed under this condition.

[0518] Crystallisation of BA.2.75 RBD

[0519] Purified BA.2.75 RBD was deglycosylated with Endoglycosidase H1 and mixed with ACE2 at a 1 :1 molar ratio, final concentration 13.0 mg ml -1 A nanoliter sitting-drop vapour-diffusion method was used to set up the initial screen of crystals using a Cartesian robot in Crystalquick 96-well X-plates (Greiner Bio-One) with 100 nL protein plus 100 nL reservoir per drop, as previously described (Walter et al., 2003). Crystals of BA.2.75 RBD-ACE2 complex were formed under Hampton Research PEG Rx condition 2-25 containing 0.1% (w / v) n-octyl-b-D-glucoside, 0.1 M trisodium citrate dihydrate pH 5.5 and 22% (w / v) PEG 3350. Diffraction data were collected at beamline I03 at Diamond Light Source, UK, at 100 K using an automated queue system that allows unattended data collection (https: / / www.diamond.ac.uk / Instruments / Mx / I03 / I03-Manual / Unattended-Data-Collections.html).

[0520] X-ray data collection, structure determination and refinement

[0521] Diffraction data were collected at beamline I03 at Diamond Light Source, UK, at 100 K. All data were collected as part of an automated queue system that allows unattended data collection (https: / / www.diamond.ac.uk / Instruments / Mx / I03 / I03-Manual / Unattended-Data-Collections.html). Crystals were pre-frozen by mounting them in loops and soaking in cryoprotectant containing 25% glycerol and 75% mother liquor for one second. Diffraction images were recorded on an Eiger2 XE 16M detector over 0.1° rotation (exposure time of 0.018 s per image, beam size of 80 x 20 pm, beam transmission of 10%, wavelength of 0.9808 A) at 100 K. ). Data were indexed, integrated and scaled using the automated data processing program Xia2-dials (Winter, 2010, Journal of applied crystallography 43, 186-190; Winter et al., 2018, Acta Crystallogr D Struct Biol 74, 85-97). 360° data were collected from a single crystal for each data set.

[0522] Structures were determined by molecular replacement with PHASER (McCoy et al., 2007, J Appl Crystallogr 40, 658-674). VhVl and ChCl domains with the greatest sequence similarity to previously determined SARS-CoV-2 RBD / Fab structures (Dejnirattisai et al., 2021, Cell 184, 2183-2200e2122; Dejnirattisai et al., 2021, Cell 184, 2939-2954e2939; Huo et al., 2020, Cell Host Microbe 28, 445-454; Liu et al., 2021, Cell 184, 4220-4236e4213; Supasa et al., 2021, Cell 184, 2201-2211e2207; Zhou et al., 2021, Cell 184, 2348-2361e2346; Zhou et al., 2020, Nature structural & molecular biology 27, 950-958) were used as search models for each current structure determination.

[0523] COOT model rebuilding (Emsley et al., 2010, Biological Crystallography 66, 486-501) and Phenix refinement (Liebschner et al., 2019, Acta Crystallogr D Struct Biol 75, 861-877) were used for all structures. Due to lower resolution, only the structure of the BA.1-RBD / O-12 / Beta-54 complex was subjected to rigid body and group B factor refinement.

[0524] Data collection and refinement statistics are given in Tables 19 and 25. Structure comparisons used SHP (Stuart et al., 1979, J Mol Biol 134, 109-142), PISA (Krissinel and Henrick, 2007, J Mol Biol 372, 774-797) to identify residues forming the RBD / Fab interface, and PyMOL (PyMOL Molecular Graphics System, Version 1.2r3pre, Schrodinger LLC) to make figures.

[0525] Example 5 Antibody structure

[0526] The structure of the BA.1 RBD / Fab Omi-12 / Fab Beta-54 ternary complex was determined to 2.3 A resolution (Table 19,

[0527] A). Although BLI experiments showed no significant competition binding between the two Fabs, a slight clash between them was observed. The high resolution structure of the uncomplexed Omi-12 fab (2.1 A resolution, Table 19) was modelled into the electron density of the complex (B, 6C). Superimposition of Fab 253 onto Fab Omi-12 showed that Q493R would clash with the H2 loop of Fab 253, whereas in Omi-12 the H2 adopts a slightly flattened structure. This structural change can be attributed to antibody maturation via a somatic mutation V53P in the heavy chain variable region of Omi-12, which forms a stacking interaction with Y489 (D). Figure 6 Figure 6 Figure 6

[0528] ​​​​​Omi-12 and antibody 253 both derive from the germline heavy chain IGHV1-58. Interestingly, like antibody 253, other antibodies described herein that derive from the germline heavy chain IGHV1-58, namely Beta-47, Beta-25, antibody 55, antibody 165, and antibody 318, also have a valine (V) at position 53 of the heavy chain variable region, i.e., a valine (V) at position 53 in SEQ ID NO: 262 (antibody 253), SEQ ID NO: 591 (Beta-47), SEQ ID NO: 461 (Beta-25), SEQ ID NO: 62 (antibody 55), SEQ ID NO: 182 (antibody 165), and SEQ ID NO: 332 (antibody 318). Position 53 in these sequences corresponds to position 58 according to IMGT numbering. Based on these data, modifying any of these antibodies by substituting the valine at position 53 with a proline (i.e., absolute numbering V53P, or V58P according to IMGT numbering) would result in an antibody effective against Omicron.

[0529] In addition, antibody AZD8895 (heavy chain variable region amino acid sequence provided in SEQ ID NO: 963, and light chain variable region amino acid sequence provided in SEQ ID NO: 965) also derives from the germline heavy chain IGHV1-58 (e.g., Dong et al., Nat Microbiol 6, 1233-1244 (2021)). AZD8895 has an isoleucine (I) at position 53 in the heavy chain variable region (corresponding to position 58 according to IMGT numbering). Based on the data herein, modifying the heavy chain variable region AZD8895 (SEQ ID NO: 963) by substituting the isoleucine at position 53 with a proline (i.e., using absolute numbering I53P, or using IMGT numbering I58P) would result in an antibody effective against Omicron.

[0530] Thus, the data indicate that modifying VH1-58 antibodies such that there is a proline at position 53 in the heavy chain variable region (corresponding to position 58 according to IMGT numbering) will make them particularly effective against Omicron.

[0531] ACE2 / BA.2.75 RBD structure

[0532] To elucidate the molecular mechanism of high affinity, the structure of BA.2.75 RBD with ACE2 was determined by crystallography (according to the method described in Example 4). As expected, the mode of binding was essentially indistinguishable from the mode of binding observed previously Figure 20A), despite significant rearrangements outside of the ACE2 footprint, where the flexible RBD 371-375 loop rearranges and part of the C-terminal 6xHis tag becomes ordered. Figure 20 B shows a close-up of the binding interface compared to the ACE2 / BA.2 RBD complex. In other complexes (with R or Q at RBD 493), K31 of ACE2 tends to be disordered, whereas in the BA.2.75 complex it is well ordered, allowing K31 to form a potential hydrogen bond with the glutamine side chain, which can increase affinity for ACE2.

[0533] Table

[0534] Table 1 - SEQ ID NOs of antibodies raised against early pandemic strains

[0535]

[0536]

[0537] Table 2 - SEQ ID NOs of antibodies raised against beta strains

[0538]

[0539]

[0540] Table 3 - SEQ ID NOs of antibodies raised against Omicron strains

[0541]

[0542]

[0543]

[0544]

[0545]

[0546]

[0547]

[0548]

[0549] Table 8 - Examples of mixed chain antibodies generated from antibodies derived from the same heavy chain isotype IGHV3-30

[0550]

[0551] Table 9 - Examples of mixed chain antibodies generated from antibodies derived from the same heavy chain isotype IGHV3-33

[0552]

[0553] Table 10 - Examples of mixed chain antibodies generated from antibodies derived from the same heavy chain isotype IGHV1-18

[0554]

[0555] Table 11 - Examples of mixed chain antibodies generated from antibodies derived from the same heavy chain isotype IGHV3-9

[0556]

[0557] Table 12 - Examples of mixed chain antibodies generated from antibodies derived from the same heavy chain isotype IGHV4-31

[0558]

[0559] Table 13 - IC50 titers of 22 Omicron SARS-CoV-2 specific human mAbs against live virus strains Victoria, Alpha, Beta, Gamma, Delta, and Omicron (BA.1). Table 14 - IC50 titers of 22 Omicron SARS-CoV-2 specific human mAbs against pseudovirus strains Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3.

[0560]

[0561]

[0562]

[0563] Table 15 - Early pandemic SARS-CoV-2 specific human mAbs and Beta SARS-CoV-2 specific human mAbs against pseudovirus strains Victoria, Omicron BA.1, Omicron BA.1.1, and Omicron BA.2. Table 16 - Commercial mAbs against pseudovirus strains Victoria, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, and Omicron BA.3.

[0564]

[0565] Table 17 - Properties of Omicron antibodies Table 18 - IC50 titers of 22 Omicron SARS-CoV-2 specific human mAbs or commercial mAbs against various SARS-CoV-2 strains.

[0566]

[0567]

[0568] Table 27. Primer sequences used to generate pseudoviruses. Related to plasmid construction and pseudotyped lentiviral particle production. Table 28. X-ray data collection and structure refinement statistics. Values in parentheses are for the highest resolution shell. Table 29. IC50 values for Omicron mAbs and commercial monoclonal antibodies

[0569]

[0570]

[0571] Table 30. Primer sequences used to generate pseudoviruses. Related to plasmid construction and pseudotyped lentiviral particle production.

[0572]

[0573]

[0574] Table 31. Table of SARS-CoV-2 lineages and genomic mutations Table 32. IC50 values for BA.1 mAbs and commercial mAbs

[0575]

[0576]

[0577]

[0578] Table 19 - X-ray data collection and structure refinement statistics for BA.1 RBD / Omi-12-Beta54 and Omi-12 Fab

[0579]

[0580] a Omi12 is glycosylated at N102 of the heavy chain.

[0581] b Values in parentheses are for the highest resolution shell.

[0582] c Only rigid body and group B factor refinement.

[0583] Table 20 - Comparison of BA.4 neutralization with BA.1, BA.1.1, BA.2, and BA.3 neutralization in pseudovirus assay

[0584]

[0585]

[0586] Table 21 - Activity of commercial antibodies against BA.4 and BA.5

[0587]

[0588] Table 22 - IC50 of BA.1 mAbs against PV BA.2.75 and BA.2+N460K

[0589]

[0590] Table 23 - IC50 of commercial mAbs against PV BA.2.75

[0591]

[0592]

[0593] Table 24

[0594]

[0595] Table 25 - X-ray data collection and structure refinement statistics for BA.2.75 RBD / ACE2

[0596]

[0597] a Values in parentheses are for the highest resolution shell.

[0598] Table 26. IC50 values for Omicron mAbs

[0599]

[0600] SEQUENCE LISTING

[0601]

[0602] Nucleotide sequences of heavy and light chain variable regions of selected antibodies

[0603]

[0604]

[0605] Amino acid sequences of CDRs

[0606] a

[0607]

[0608] b

[0609]

[0610] Amino acid sequences of CDRs

[0611]

[0612] Amino acid sequences of CDRs

[0613]

[0614]

[0615] ​

[0616] a

[0617]

[0618] b

[0619]

[0620] ​

[0621] Amino acid sequences of heavy and light chain variable regions of selected antibodies

[0622]

[0623]

[0624]

[0625]

[0626]

[0627]

[0628] Nucleotide sequences of heavy and light chain variable regions of selected antibodies

[0629]

[0630]

[0631]

[0632]

[0633]

[0634]

[0635]

[0636]

[0637]

[0638]

[0639]

[0640]

[0641]

[0642] Amino acid sequences of CDRs

[0643]

[0644]

[0645]

[0646]

[0647]

[0648] Amino acid sequences of heavy and light chain variable regions of selected antibodies

[0649]

[0650]

[0651]

[0652]

[0653]

[0654]

[0655] ​

[0656]

[0657]

[0658]

[0659]

[0660]

[0661]

[0662]

[0663]

[0664]

[0665] ​

[0666]

[0667]

[0668]

[0669]

[0670] Amino acid sequences of heavy and light chain variable regions of selected antibodies

[0671]

[0672]

[0673]

[0674]

[0675]

[0676]

[0677] Nucleotide sequences of heavy and light chain variable regions of selected antibodies

[0678]

[0679]

[0680]

[0681]

[0682]

[0683]

[0684]

[0685]

[0686]

[0687] ​

[0688]

[0689]

[0690]

[0691]

[0692] SEQ ID NO: 961 - Amino acid sequence encoded by IGHV1-58 germline V gene sequence MQLVQSGPEVKKPGTSVKVSCKASGFTFTSSAVQWVRQARGQRLEWIGWIVVGSGNTNYAQKFQERVTITRDMSTSTAYMELSSLRSEDTAVYYCAA

[0693] SEQ ID NO: 962 - AZD8895 (COV2-2196) heavy chain variable region nucleotide sequence

[0694]

[0695] SEQ ID NO: 963 - AZD8895 (COV2-2196) heavy chain variable region amino acid sequence:

[0696] [0...

Claims

1. An antibody capable of binding to the spike protein of the coronavirus SARS-CoV-2, wherein the antibody comprises: (a) a heavy chain variable domain sequence as set forth in SEQ ID NO: 952; and (b) a light chain variable domain sequence as set forth in SEQ ID NO:

954.

2. The antibody of claim 1, wherein the antibody comprises an Fc region.

3. The antibody of claim 2, wherein the antibody comprises an IgGl constant region.

4. The antibody of claim 2 or 3, wherein the Fc region comprises at least one modification such that serum half-life is prolonged.

5. The antibody of claim 4, wherein the Fc region comprises the mutations M252Y / S254T / T256E (YTE).

6. The antibody of claim 1, wherein the antibody is a single chain antibody, a Fab, a modified Fab, a Fab', a modified Fab', a F(ab')2, a Fv, a Fab-Fv, a Fab-dsFv, or a scFv.

7. A polynucleotide encoding the light chain variable domain and the heavy chain variable domain of the antibody of any one of claims 1 to 6.

8. A vector comprising one or more polynucleotides of claim 7.

9. A host cell comprising one or more polynucleotides of claim 7 or one or more vectors of claim 8.

10. A method for producing an antibody capable of binding to the spike protein of the coronavirus SARS-CoV-2, the method comprising culturing the host cell of claim 9 and isolating the antibody from the culture.

11. A pharmaceutical composition comprising: (a) the antibody of any one of claims 1 to 6 (b) at least one pharmaceutically acceptable diluent or carrier.

12. The antibody of any one of claims 1 to 6 or the pharmaceutical composition of claim 11 for use in a method of treatment of the human or animal body by therapy.

13. The antibody of any one of claims 1 to 6 or the pharmaceutical composition of claim 11 for use in a method of treating a disease or complication associated with SARS-CoV-2 infection.

14. The antibody for use of claim 13, wherein the disease or complication associated with SARS-CoV-2 infection is COVID-19.

15. The antibody for use of claim 13, wherein the SARS-CoV-2 infection is caused by a SARS-CoV-2 strain of lineage alpha, beta, gamma, delta, or omicron.

16. The antibody for use according to claim 15, wherein the lineage of the omicron strain is Omicron BA.2.11, Omicron BA.2.12.1, Omicron BA.2.13, Omicron BA.2.3.20, Omicron BA.2.10.4, Omicron BA.1, Omicron BA.1.1, Omicron BA.2, Omicron BA.2.75, BA.2.75.2, Omicron BA.3, Omicron BA.4.6, Omicron BA.4 / 5, Omicron BJ.1, Omicron BS.1, Omicron BN.1, Omicron BF.7, Omicron BQ.1, Omicron BQ.1.1 Omicron XBB and / or Omicron XBB.1.

Images