The current coronavirus disease 2019 (COVID-19) pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has cost the global economy dearly, besides the enormous loss of life and health. In the absence of effective antivirals, vaccines and therapeutic antibodies have been a major area of research. A new preprint on the bioRxiv* server reports highly protective neutralizing antibodies against the virus, directed against different epitopes than the majority of monoclonal antibodies reported in current literature.
The virus mediates infection of the host cell by initiating attachment to the host receptor, the angiotensin-converting enzyme 2 (ACE2), by its receptor-binding domain (RBD) on the characteristic spike protein.
The spike is made up of two subunits, the S1 that contains the RBD for receptor engagement, and the S2 for viral-cell membrane fusion and virus internalization. The S1 itself contains the N-terminal domain (NTD) and the C-terminal domain (CTD). The latter is the RBD.
Some researchers suggest that receptors of the C-type lectin superfamily, namely, L-SIGN and DC-SIGN, are also receptors for SARS-CoV-2 on human cells, binding to the viral S1 NTD. If so, these alternative receptors could explain the virus's ability to infect even tissues that have low or undetectable ACE2 expression.
The spike is thus the target of 90% of currently identified therapeutic antibodies, which disrupt the binding of the virus with ACE2 to achieve their neutralizing effects. The chief identification criterion is thus the ability of the antibody to bind the spike RBD.
The rapid emergence and rise to dominance of more infective variants such as the UK, South African and Brazil variants highlight the inadequacy of this approach. The extensive use of therapeutic antibodies promotes the emergence of escape mutations, especially when a single monoclonal antibody targets one region of the virus, exerting selective pressure.
The most effective way to avoid this is to use multiple monoclonals targeting a range of epitopes on the virus. Such antibody cocktails are demonstrably efficient at preventing antibody resistance in HIV, and with the current virus as well. The development of such neutralizing antibodies is thus of great importance.
Coronavirus-related research suggests that the NTD may also play a role in host cell infection, since it shares loop region sequences with MERS-CoV. These loops are capable of binding sialosides, which are used by the latter virus as receptors to infect the host cells in a range of tissues.
The researchers screened antibodies from a phage-display library, isolated from patients with severe COVID-19. The blood samples used all had anti-spike antibodies at titers above 10,000, and neutralization potency above 1,000, and were pooled to construct the library.
This led to the identification of 12 antibodies, mostly from the IGHV1-24 germline family. The enrichment of this family in anti-SARS-CoV-2 antibodies and in IgG-producing B cell clones from COVID-19 patients indicates that it is preferentially involved in B cell binding to viral epitopes.
One of the antibodies, BLN14, showed sequences from three similarly enriched families, which is interesting since "the combination of VH and VL was the result of combinatorial pairing of heavy and light chains during library construction," though not present in the humoral immune response of the patients themselves. In other words, it is possible that these families, having been positively selected for in the screen, are combined because they produce high-affinity antibodies to the virus.
These antibodies have a long CDHR3 loop like other non-RBD neutralizing antibodies against this virus, as well as low somatic hypermutations in the IGHV germline, as seen in many antibodies elicited by SARS-CoV-2 and in IgG-secreting B cell clones from COVID-19 patients.
Affinity and specificity
The researchers found that these anti-NTD antibodies are very specific in their epitope binding, and recognize both the full-length spike and the S1 subunit in so doing. The binding affinities of the 12 antibodies ranged from 0.7 nM to 34 nM. The highest affinity was with BLN1 and BLN12, and the lowest with BLN4.
The neutralizing capacity by plaque reduction neutralization test (PRNT) showed that all of the 12 antibodies were powerful inhibitors of viral infection, with 50% inhibitory values (IC50) between just 0.008 μg/ml (highly potent) and 54.9 μg/ml.
The researchers found that the 12 anti-NTD antibodies belonged to three groups recognizing three different epitopes, the first containing only BLN8, the second BLN4 and BLN6, and all the rest in group III. Nine of the group III antibodies had high neutralization potency. This suggests that of the three epitopes, this group targets a very vulnerable region of the virus.
The study also shows that several of these antibodies recognize peptide epitopes that are N-glycosylation sites as well. The subsequent examination of the antibody groups against a glycan microarray showed a strong preference for N-glycan-linked structures.
Replacement of the sialic acid at the terminus by N-glycolylneuraminic acid abolished binding. Further study showed that for some antibodies at least, N-glycans are essential for binding, being involved in epitope recognition along with the amino acid sequences.
Using a hACE2 binding inhibition assay, the researchers explored the mechanism of neutralization of anti-NTD antibodies. They found, expectedly, that none of the antibodies prevented spike-hACE2 binding, confirming that the RBD is not targeted by these antibodies.
However, when they examined NTD-L-SIGN binding, they found significant interaction. Following this, they observed partial inhibition, ranging from 25% to 50%, of NTD-L-SIGN binding, by the selected antibodies. The mechanism of neutralization is possibly partially mediated by disrupting this interaction, either directly, or perhaps because these antibodies stabilize an unfavorable conformational state of viral fusion proteins.
If C-type lectins serve as alternative entry receptors for this virus, "it may be speculated that anti-NTD mAbs may also exert their therapeutic activity by limiting the spreading of the virus in the body."
Using a lethal mouse model (K18-hACE2 transgenic mice), they tested the in vivo protective efficacy of selected anti-NTD antibodies. They constructed full recombinant IgG molecules expressing two of the anti-NTD antibodies with high in vitro neutralization potencies, namely, BLN12 and BLN14.
The mice were first infected with a dose sufficient to cause over 75% of them to die within 7-10 days from infection. At two days post-infection, a single dose (either 0.1 or 0.01 mg/mouse) of one of these antibodies was administered. They found that the treated mice were completely protected at the 0.1 mg dose, and 80-85% of mice at the lower dose. In the second group, survivors recovered completely by day 10 post-infection.
These results indicate a high efficacy of both mAbs in vivo, even at low doses and at a late time point post infection."
Endogenous antibody response
The antibody-treated mice were also found to show robust neutralizing antibody responses when treated with two antibody doses post-infection. This is a significant advantage of such treatment, especially since re-infection has been reported to occur with SARS-CoV-2.
What are the implications?
The unusually high neutralizing capacity of two anti-NTD antibodies, with complete protection of lethally infected mice even when given late in infection, promises much for their development as therapeutic antibodies. "In combination with anti-RBD mAbs, therapeutic administration of these antibodies may circumvent complications related to the occurrence of viral escape mutants."
More study is required as to how the mice, which were infected via the hACE2 receptor, could be fully protected by these antibodies, which do not prevent spike-ACE2 binding. Nonetheless, BLN12 and BLN14 monoclonals are excellent therapeutic candidates for SARS-CoV-2 infection.
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.