The ongoing coronavirus disease 2019 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is only one of several very similar zoonotic diseases that have led to outbreaks over the last two decades. A neutralizing response aimed at a common viral target rather than proteins specific to each virus would be very helpful in ending the current pandemic as well as forestalling future similar outbreaks.
A new study, released as a preprint on the bioRxiv* server, reports on a set of antibodies from COVID-19 donor serum, which seems to be cross-reactive across multiple Sarbecoviruses, including the very similar SARS-CoV-2, SARS-CoV and Middle East Respiratory Syndrome (MERS)-CoV.
The spike antigen
All these coronaviruses (CoVs) have trimeric spike glycoproteins on their surface that mediate virus-receptor attachment and viral entry into the host cell. The spike has a receptor-binding domain (RBD) at the apex of each protomer.
The angiotensin-converting enzyme 2 (ACE2) is the host cell receptor for SARS-CoV and SARS-CoV-2, as well as some of the other animal and seasonal human CoVs. The RBD exists in either the ‘up’ or ‘down’ conformation, but is accessible to the receptor only in the former.
The CoV spike protein RBDs are related to each other and appear to branch from a common stem. Their importance to successful viral entry makes them a natural and appropriate target for neutralizing antibodies.
They have been classified into four categories: class 1 is encoded by VH3-53/VH3-63 germline genes, and their short heavy chain complementarity determining region 3 (CDRH3) binds at an epitope that overlaps the ACE2 binding site and can bind only if the RBD is ‘up.’
Class 2 has a similarly overlapping epitope, but one that can bind both ‘up’ and ‘down’ RBDs. Class 3 binds to ‘up’ and ‘down’ RBDs on opposite sides, near an N-glycosylation site. Class 4 antibodies are typically weakly neutralizing and bind to a cryptic inward-directed epitope on ‘up’ RBDs.
The most potent neutralizing antibodies specific to SARS-CoV-2 are either class 1 or 2, and act by blocking RBD-ACE2 binding. They are therefore less conserved across Sarbecoviruses.
Class 3 and 4 antibodies directed against the RBD are better choices for broadly neutralizing activity, as shown by S309, a class 3 antibody to SARS-CoV-2 isolated from a SARS-CoV-infected donor.
Thus, the potency of neutralization is often traded off for increased cross-reactivity, as is common with circulating viruses that have multiple variants.
Broad binding spectrum
The current study focused on C118 and C022, both class 4 antibodies from COVID-19-positive donors, which were able to bind to and neutralize most of the Sarbecovirus RBDs from an array of 12, belonging to different clades, as well as SARS-CoV-2 variants of concern (VOCs) which have immune-evading and/or higher infectivity characteristics.
Only two RBDs were not recognized by C022, while C118 bound to all of them. To put this in perspective, class 3 antibody S309 bound half of them. When tested against the variant RBDs from circulating SARS-CoV-2, including that of B.220.127.116.11 and B.1.351 (the UK and South African VOCs), both showed comparable binding. This property is shared by the class 4 antibody CR3022 as well as S309.
The crystal structure of C118 bound to SARS-CoV, and of C022 in complex with SARS-CoV-2, showed a highly conserved RBD epitope. This has been observed to be bound by other cross-reactive but less neutralizing SARS-CoV-2 class 4 antibodies such as CR3022, S304/S2A4 and EY6A.
The interactions are numerous and are based on van der Waals and polar bonds with the RBD. The Fab (antigen-binding fragment) would cause steric hindrance to ACE2-RBD binding, as well as ACE2 binding to the antibody-RBD complex. Thus, C118, C022, and C144 appear to compete with ACE2 for binding to the RBD, but not CR3022.
This suggests “a primary neutralization mechanism for C022 and C118 that prevents spike attachment to host cell ACE2 receptors.”
When tested against a set of pseudoviruses bearing mutant SARS-CoV-2 spikes with a single amino acid substitution, both C118 and C022 showed equal potency of neutralization as with the wildtype virus bearing the D614G spike.
The 50% inhibitory concentrations (IC50 values) were <1 μg/mL, with both antibodies, for almost all the viruses tested. The exception was C118 neutralization of SARS-CoV spike, with an IC50 almost five times as high as for the rest. In comparison, CR3022 is poorly neutralizing except for SARS-CoV and WIV1.
S309 is a class 3 antibody but showed high neutralization potency at nanomolar concentrations, except for B.18.104.22.168 and SHC014.
In addition to their broad spectrum, both C118 and C022 have greater neutralizing ability because they block part of the ACE2 binding site.
Highly conserved sequences
Most residues are highly conserved or substituted by synonymous residues between SARS-CoV-2 and other RBDs, which probably accounts for the broad cross-reactivity. In fact, 70% of residues are shared by both antibodies, and the majority of RBD contacts were completely identical across Sarbecoviruses.
Backbone interactions provide double protection
Both these antibodies have dominant long CDRH3 loops that allow binding to a highly conserved set of residues common to all Sarbecoviruses in an upward-facing direction that allows it to overlap the ACE2 binding site.
The long CDRH3 loops also interact extensively with the RBD backbone, especially with the cryptic RBD epitope at its base. Long backbone interactions allow binding to occur even with substituted residues in the side chains in different sarbecovirus strains or SARS-CoV-2 variants.
This type of substitution is rare, however, with the virus strains included in this study. However, this double layer of protection against escape mutations could explain the broad cross-reactivity and neutralization ability across Sarbecoviruses.
Increased intra-spike crosslinking may improve avidity
Class 4 antibodies not only require two RBDs in the ‘up’ position but also RBD rotation to overcome steric hindrance from adjacent RBDs, in some cases, as with S309.
When examined by cryo-electron microscopy, the spike trimer-C118 complex revealed a greater distance between the RBDs in the trimer compared to complexes between the Fab of antibodies from classes 1 to 3 and the spike.
The separation was also greater than seen in the ACE2-spike complexes. The importance of the greater distances available between the spike RBDs in the receptor-bound form is the possibility they offer for crosslinking within each spike, which is associated with increased avidity of binding.
This avidity effect was not observed with C118 for SARS-CoV-2 and SARS-CoV neutralization, however, though it is present with WIV1 and SHC014. Conversely, it is strong in C022 for all four pseudoviruses.
Moreover, the findings suggest that binding is followed by loss of spike trimer stability, which may contribute to neutralization potency.
What are the implications?
These results define a cross-reactive class 4-like epitope on sarbecovirus RBDs that can be targeted in vaccine design and illustrate a mechanism by which the cryptic RBD epitope can be accessed on intact CoV S trimers.”
These antibodies showed potent neutralization activity against all the sarbecovirus and SARS-CoV-2 viruses tested here, remaining robustly neutralizing against natural RBD mutations which define several VOCs in current circulation.
Mutations tend to accumulate on the ACE2-bindng region of the RBD, thus helping the virus to evade specific RBD-targeting antibodies. These antibodies possess multiple mechanisms to reduce the impact of such mutations.
The class 4 antibodies described here, C118 and C022, are of the type that results from natural infection, and thus their binding sites should be studied to come up with an effective immunogen design.
Further work may focus on proper selection of potent neutralizing class 4 RBD-targeting antibodies for increased potency, for their use in the treatment of this condition, as well as to design better and broader immunogens in the future.
bioRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.