Scientists at Fred Hutchinson Cancer Research Center in Seattle have shown that a potent antibody from a COVID-19 survivor interferes with a key feature on the surface of the coronavirus’s distinctive spikes and induces critical pieces of those spikes to break off in the process.
The antibody – a tiny, Y-shaped protein that is one of the body’s premier weapons against pathogens including viruses – was isolated by the Fred Hutch team from a blood sample received from a Washington state patient in the early days of the pandemic.
The team led by Drs. Leo Stamatatos, Andrew McGuire and Marie Pancera previously reported that, among dozens of different antibodies generated naturally by the patient, this one – dubbed CV30 – was 530 times more potent than any of its competitors.
Using tools derived from high-energy physics, Hutch structural biologist Pancera and her postdoctoral fellow Dr. Nicholas Hurlburt have now mapped the molecular structure of CV30. They and their colleagues published their results online today in the journal Nature Communications.
The product of their research is a set of computer-generated 3-D images that look to the untrained eye as an unruly mass of noodles.
But to scientists they show the precise shapes of proteins comprising critical surface structures of antibodies, the coronavirus spike and the spike’s binding site on human cells. The models depict how these structures can fit together like pieces of a 3-D puzzle.
“Our study shows that this antibody neutralizes the virus with two mechanisms. One is that it overlaps the virus’s target site on human cells, the other is that it induces shedding or dissociation of part of the spike from the rest,” Pancera said.
On the surface of the complex structure of the antibody is a spot on the tips of each of its floppy, Y-shaped arms.
This infinitesimally small patch of molecules can neatly stretch across a spot on the coronavirus spike, a site that otherwise works like a grappling hook to grab onto a docking site on human cells.
The target for those hooks is the ACE2 receptor, a protein found on the surfaces of cells that line human lung tissues and blood vessels.
But if CV30 antibodies cover those hooks, the coronavirus cannot dock easily with the ACE2 receptor. Its ability to infect cells is blunted.
This very effective antibody not only jams the business end of the coronavirus spike, it apparently causes a section of that spike, known as S1, to shear off. Hutch researcher McGuire and his laboratory team performed an experiment showing that, in the presence of this antibody, there is reduction of antibody binding over time, suggesting the S1 section was shed from the spike surface.
The S1 protein plays a crucial role in helping the coronavirus to enter cells. Research indicates that after the spike makes initial contact with the ACE2 receptor, the S1 protein swings like a gate to help the virus fuse with the captured cell surface and slip inside.
Once within a cell, the virus hijacks components of its gene and protein-making machinery to make multiple copies of itself that are ultimately released to infect other target cells.
Fred Hutch structural biologists developed 3-D images of an antibody fished from the blood of an early COVID-19 survivor that efficiently neutralized the coronavirus. Dr. Nicholas Hurlburt, who helped develop the images, narrates this short video showing how that antibody interacts with the notorious spikes of the coronavirus, blocking their ability to bind to a receptor on human cells that otherwise presents a doorway to infection. Credit: Video by Robert Hood / Fred Hutch News Service, Images courtesy Nicholas Hurlburt.
The incredibly small size of antibodies is difficult to comprehend. These proteins are so small they would appear to swarm like mosquitos around a virus whose structure can only be seen using the most powerful of microscopes. The tiny molecular features Pancera’s team focused on the tips of the antibody protein are measured in nanometers—billionths of a meter.
Yet structural biologists equipped with the right tools can now build accurate 3-D images of these proteins, deduce how parts of these structures fit like puzzle pieces, and even animate their interactions.
Key to building models of these nanoscale proteins is the use of X-ray crystallography. Structural biologists determine the shapes of proteins by illuminating frozen, crystalized samples of these molecules with extremely powerful X-rays.
The most powerful X-rays come from a gigantic instrument known as a synchrotron light source. Born from atom-smashing experiments dating back to the 1930s, a synchrotron is a ring of massively powerful magnets that are used to accelerate a stream of electrons around a circular track at close to the speed of light. Synchrotrons are so costly that only governments can build and operate them. There are only 40 of them in the world.
Pancera’s work used the Advanced Photon Source, a synchrotron at Argonne National Laboratory near Chicago, which is run by the University of Chicago and the U.S. Department of Energy. Argonne’s ring is 1,200 feet in diameter and sits on an 80-acre site.
As the electrons whiz around the synchrotron ring, they give off enormously powerful X-rays—far brighter than the sun but delivered in flashes of beams smaller than a pinpoint.
Structural biologists from around the world rely on these brilliant X-ray beamlines to illuminate frozen crystals of proteins. They reveal their structure in the way these bright beams are bent as they pass though the molecules. It takes powerful computers to translate the data readout from these synchrotron experiments into the images of proteins that are eventually completed by structural biologists.
The Fred Hutch team’s work on CV30 builds on that of other structural biologists who are studying a growing family of potent neutralizing antibodies against the coronavirus.
The goal of most coronavirus vaccine candidates is to stimulate and train the immune system to make similar neutralizing antibodies, which can recognize the virus as an invader and stop COVID-19 infections before they can take hold.
Neutralizing antibodies from the blood of recovered COVID-19 patients may also be infused into infected patients—an experimental approach known as convalescent plasma therapy.
The donated plasma contains a wide variety of different antibodies of varying potency. Although once thought promising, recent studies have cast doubt on its effectiveness.
However, pharmaceutical companies are experimenting with combinations of potent neutralizing antibodies that can be grown in a laboratory. These “monoclonal antibody cocktails” can be produced at industrial scale for delivery by infusion to infected patients or given as prophylactic drugs to prevent infection.
After coming down with COVID-19, President Trump received an experimental monoclonal antibody drug being tested in clinical trials by the biotech company Regeneron, and he attributes his apparently quick recovery to the advanced medical treatment he received.
The Fred Hutch research team holds out hope that the protein they discovered, CV30, may prove to be useful in the prevention or treatment of COVID-19. To find out, this antibody, along with other candidate proteins their team is studying, need to be tested preclinically and then in human trials.
“It is too early to tell how good they might be,” Pancera said.
COVID-19 was declared a pandemic in March 2020 by the World Health Organization1. As of June 11th, 2020, there were ~ 7.4 M infections and over 415,000 deaths worldwide2. It is caused by a coronavirus of the beta family, named SARS-CoV-23, as it is closely related to SARS-CoV4.
Their genomes share 80% identity and they utilize angiotensin-converting enzyme 2 (ACE2) as receptor for entry5–11. Viral entry depends on the SARS-CoV-2 spike glycoprotein, a class I fusion protein comprised of two subunits, S1 and S2. S1 mediates ACE2 binding through the receptor binding domain (RBD), while the S2 subunit mediates fusion.
Overall the spike shares 76% amino acid sequence homology with SARS4. High resolutions structures of the SARS-CoV-2 stabilized spike in the prefusion revealed that the RBD can be seen in a ‘up’ or ‘down’ conformation5,6.It’s been shown that some of the neutralizing antibodies bind the RBD in the ‘up’ conformation similar to when the ACE2 receptor binds12.
Currently there is no vaccine available to prevent SARS-CoV-2 infection and highly effective therapeutics have not been developed yet either. The host immune response to this new coronavirus is also not well understood.
We, and others, sought to characterize the humoral immune response from infected COVID-19 patients12–14. Recently, we isolated a neutralizing antibody, named CV30, which binds the receptor binding domain (RBD), neutralizes with 0.03 μg/ml and competes binding with ACE215.
However, the molecular mechanism by which CV30 blocked ACE2 binding was unknown. Herein, we present the 2.75 A crystal structure of SARS-CoV-2 RBD in complex with the Fab of CV30 (Extended Data Table 1).
CV30 binds almost exclusively to the concave ACE2 binding epitope (also known as the receptor binding motif (RBM)) of the RBD using all six CDR loops with a total buried surface area of ~1004 Å2, ~750 Å2 from the heavy chain and ~254 Å2 from the kappa chain (Fig. 1A).
20 residues from heavy chains and 10 residues from the kappa chain interact with the RBD, forming 13 and 2 hydrogen bonds, respectively (Fig. 1C and Extended Data Table 2). There are 29 residues from the RBD that interact with CV30, 19 residues with the heavy chain, 7 residues with the light chain, and 3 residues with both (Extended Data Table 2).
Of the 29 interacting residues from the SARS-CoV-2 RBD, only 16 are conserved in the SARS-CoV S protein RBD (Fig. 2c), which could explain the lack of cross-reactivity of CV30 to SARS-CoV S15.
The CV30 heavy chain is minimally mutated with only a two-residue change from the germline and both of these residues (Val27-Ile28) are located in the CDRH1 and form nonpolar interactions with the RBD.
We reverted these residues to germline to assess their role. Interestingly, the germline CV30 (glCV30) antibody bound to RBD with ~100-fold lower affinity (407 nM affinity) (Fig 1d and Extended Data Table 3) compared to CV30 (3.6 nM15) with a very large difference in the off-rate. glCV30 neutralized SARS-CoV-2 with ~500-fold difference with an IC50 of 16.5 vs 0.03 μg/mL for CV30 (Fig. 1e).
Val27 forms a weak non-polar interaction with the RBD Asn487 and sits in a pocket formed by CDRH1 and 3. Although it is unclear, Phe27 presents in glCV30 could change the electrostatic environment. The Ile28 sidechain forms non-polar interactions with the RBD Gly476-Ser447, particularly the Cγ atom, which the glCV30 Thr would be incapable of making. Thus, minimal affinity maturation of CV30 significantly impacted the ability of this mAb to neutralize SARS-CoV-2.
CV30 competes with ACE2 for binding to the RBD15 and we therefore examined the structural mechanism of the receptor blocking by superimposing the SARS-CoV-2 RBD/ACE2 complex (PDB: 6LZG)9 with the CV30 Fab/RBD complex. The structure of the RBD was used to align the two complexes and showed that CV30 binding did not induce any conformational changes in the RBD from the ACE2-bound complex.
The aligned RBD had a RMSD of 0.353 Å over 166 Cα atoms. The structure reveals that the CV30 epitope overlaps almost completely with the ACE2 epitope. A total of 26 residues of the SARS-CoV-2 RBD interact with hACE2, CV30 binds to 19 of these residues (Fig. 2A), indicating that CV30 neutralizes the virus by preventing the binding of ACE2 to RBD by direct steric interactions.
Recently, the structure of two potent neutralizing anti-RBD antibodies were published, B38 and CB612,14. CV30 shares a similar germline heavy chain V-genes but all three have diverse germline kappa V-genes (CV30 is IGKV3–2001, B38 is IGKV1–901, CB6 is IGKV1–3901, Extended Data Fig. 1). Both CV30 and B38 use IGHV3–5301 while CB6 uses IGHV3–6601, which is only one amino acid different than 3–5301 (Val12 which does not make contact with the epitope).
CV30 and CB6 each have higher affinities, 3.6 nM and 2.5 nM, respectively, than B38, 70.1 nM12,14,15. Differences in affinity translate into differences in neutralization potency (the IC50s for CV30 and CB6 are 0.03 and 0.036 μg/mL, respectively, and that of B38 is 0.177 μg/mL).
Interestingly, Thr28 was also mutated from germline to Ile in B38 but Phe27 was not. CB6 lacks both mutations found in CV30.
Differences in other regions of the antibody, such as the CDRH3 and light chain are likely responsible for the overall potency all these antibodies (see below). To investigate the binding mechanism of the three antibodies, a superposition of the structures was created. All three bind in a nearly identical manner with the same angle of approach and similar footprints (Fig. 2b).
The alignment of the Fv regions of B38 and CB6 to the Fv region of CV30 had a RMSD of 0.240Å over 100 Cα atoms and 0.329Å over 98 Cα atoms, respectively. Mapping the binding interactions of the RBD to each of the antibodies reveals a close overlap in the binding mechanism (Fig. 2c–d).
The footprint of the heavy chain is nearly identical, as expected from the shared germline V-gene and sequence similarity. CV30 and CB6 both have longer CDRH3 and bind with higher buried surface area, ~263 and ~251 Å2, respectively, than B38 (~203 Å2) (Fig. 2d, Extended Data Fig. 1).
The large difference is in the light chain. CV30 has the smallest binding interaction at ~254 Å2, B38 has the largest interaction at ~497 Å2 and then CB6 at ~354Å2. One of the more interesting findings was the interaction of Thr56 in the CV30 CDRK2 which reaches across the RBD and interacts Phe486, an interaction that is not found in the other two antibodies (Extended Data Fig. 1).
In conclusion, our structure indicates that potent neutralizing antibodies against SARS-CoV-2 bind the receptor binding motif in the RBD, overlapping the ACE2 binding site, but recognize residues that are specific for SARS-CoV-2 only, thus explaining the lack of cross neutralization with SARS-CoV.
It is noteworthy that potently neutralizing antibodies isolated from multiple individuals use the same or similar VH gene to target their epitope. Additionally, the minimal affinity maturation observed 21 days after infection in the VH gene of CV30 showed ~100–500-fold increase in affinity and neutralization potency, indicating that further affinity maturation may increase potency and potential cross-reactivity.
Our studies indicate that the RBD is a promising target for vaccine design and that these potently neutralizing antibodies should be explored as a treatment for COVID-19 infection.
- 1. Mahase E. Covid-19: WHO declares pandemic because of “alarming levels” of spread, severity, and inaction. BMJ 368, m1036 (2020). [PubMed] [Google Scholar]
- 2. Dong E., Du H. & Gardner L. An interactive web-based dashboard to track COVID-19 in real time. Lancet Infect Dis (2020). [PMC free article] [PubMed] [Google Scholar]
- 3. Wu Y. et al. SARS-CoV-2 is an appropriate name for the new coronavirus. Lancet (2020). [PMC free article] [PubMed] [Google Scholar]
- 4. Wan Y., Shang J., Graham R., Baric R.S. & Li F. Receptor Recognition by the Novel Coronavirus from Wuhan: an Analysis Based on Decade-Long Structural Studies of SARS Coronavirus. J Virol 94(2020). [PMC free article] [PubMed] [Google Scholar]
- 5. Wrapp D. et al. Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation. Science 367, 1260–1263 (2020). [PMC free article] [PubMed] [Google Scholar]
- 6. Walls A.C. et al. Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein. Cell (2020). [PMC free article] [PubMed] [Google Scholar]
- 7. Hoffmann M. et al. SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor. Cell (2020). [PMC free article] [PubMed] [Google Scholar]
- 8. Yan R. et al. Structural basis for the recognition of SARS-CoV-2 by full-length human ACE2. Science 367, 1444–1448 (2020). [PMC free article] [PubMed] [Google Scholar]
- 9. Wang Q. et al. Structural and Functional Basis of SARS-CoV-2 Entry by Using Human ACE2. Cell (2020). [PMC free article] [PubMed] [Google Scholar]
- 10. Letko M., Marzi A. & Munster V. Functional assessment of cell entry and receptor usage for SARS-CoV-2 and other lineage B betacoronaviruses. Nat Microbiol 5, 562–569 (2020). [PMC free article] [PubMed] [Google Scholar]
- 11. Ou X. et al. Characterization of spike glycoprotein of SARS-CoV-2 on virus entry and its immune cross-reactivity with SARS-CoV. Nat Commun 11, 1620 (2020). [PMC free article] [PubMed] [Google Scholar]
- 12. Shi R. et al. A human neutralizing antibody targets the receptor binding site of SARS-CoV-2. Nature (2020). [PubMed] [Google Scholar]
- 13. Ju B. et al. Human neutralizing antibodies elicited by SARS-CoV-2 infection. Nature (2020). [PubMed] [Google Scholar]
- 14. Wu Y. et al. A noncompeting pair of human neutralizing antibodies block COVID-19 virus binding to its receptor ACE2. Science (2020). [PMC free article] [PubMed] [Google Scholar]
- 15. Seydoux E. et al. Analysis of a SARS-CoV-2 infected individual reveals development of potent neutralizing antibodies to distinct epitopes with limited somatic mutation. Immunity (2020). [PMC free article] [PubMed] [Google Scholar]
- 16. Meuris L. et al. GlycoDelete engineering of mammalian cells simplifies N-glycosylation of recombinant proteins. Nat Biotechnol 32, 485–9 (2014). [PMC free article] [PubMed] [Google Scholar]
- 17. Kabsch W. Xds. Acta Crystallogr D Biol Crystallogr 66, 125–32 (2010). [PMC free article] [PubMed] [Google Scholar]
- 18. McCoy A.J. et al. Phaser crystallographic software. J Appl Crystallogr 40, 658–674 (2007). [PMC free article] [PubMed] [Google Scholar]
- 19. Teplyakov A. et al. Structural diversity in a human antibody germline library. MAbs 8, 1045–63 (2016). [PMC free article] [PubMed] [Google Scholar]
- 20. Emsley P. & Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60, 2126–32 (2004). [PubMed] [Google Scholar]
- 21. Adams P.D. et al. Recent developments in the PHENIX software for automated crystallographic structure determination. J Synchrotron Radiat 11, 53–5 (2004). [PubMed] [Google Scholar]
- 22. Crawford K.H.D. et al. Protocol and Reagents for Pseudotyping Lentiviral Particles with SARS CoV-2 Spike Protein for Neutralization Assays. Viruses 12(2020). [PMC free article] [PubMed] [Google Scholar]
More information: Nicholas K. Hurlburt et al, Structural basis for potent neutralization of SARS-CoV-2 and role of antibody affinity maturation, Nature Communications (2020). DOI: 10.1038/s41467-020-19231-9