key: cord-354868-pqn59ojj
authors: Yao, Hebang; Cai, Hongmin; Li, Tingting; Zhou, Bingjie; Qin, Wenming; Lavillette, Dimitri; Li, Dianfan
title: A high-affinity RBD-targeting nanobody improves fusion partner’s potency against SARS-CoV-2
date: 2020-09-25
journal: bioRxiv
DOI: 10.1101/2020.09.24.312595
sha: 
doc_id: 354868
cord_uid: pqn59ojj

A key step to the SARS-CoV-2 infection is the attachment of its Spike receptor-binding domain (S RBD) to the host receptor ACE2. Considerable research have been devoted to the development of neutralizing antibodies, including llama-derived single-chain nanobodies, to target the receptor-binding motif (RBM) and to block ACE2-RBD binding. Simple and effective strategies to increase potency are desirable for such studies when antibodies are only modestly effective. Here, we identify and characterize a high-affinity synthetic nanobody (sybody, SR31) as a fusion partner to improve the potency of RBM-antibodies. Crystallographic studies reveal that SR31 binds to RBD at a conserved and ‘greasy’ site distal to RBM. Although SR31 distorts RBD at the interface, it does not perturb the RBM conformation, hence displaying no neutralizing activities itself. However, fusing SR31 to two modestly neutralizing sybodies dramatically increases their affinity for RBD and neutralization activity against SARS-CoV-2 pseudovirus. Our work presents a tool protein and an efficient strategy to improve nanobody potency.

SARS-CoV-2, the pathogenic virus for COVID-19, has caused a global pandemic since its first report in early December 2019 in Wuhan China (1), posing a gravely crisis for health and economic and social order. SARS-CoV-2 is heavily decorated by its surface Spike (S) (2, 3) , a single-pass membrane protein that is key for the host-virus interactions. During the infection, S is cleaved by host proteases (4, 5) , yielding the Nterminal S1 and the C-terminal S2 subunit. S1 binds to angiotensin-converting enzyme 45 2 (ACE2) (6-10) on the host cell membrane via its receptor-binding domain (RBD), causing conformational changes that trigger a secondary cleavage needed for the S2mediated membrane fusion at the plasma membrane or in the endosome. Because of this essential role, RBD has been a hot spot for the development of therapeutic monoclonal antibodies (mAbs) and vaccine (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) . 50 Llama-derived heavy chain-only antibodies (nanobodies) are attractive biotherapeutics (29) . These small (~14 kDa) proteins are robust, straightforward to produce, and amenable to engineering such as mutation and fusion. Owing to their ultra-stability, nanobodies have been reported to survive nebulization, a feature that has been explored 55 for the development of inhaled nanobodies to treat respiratory viral diseases (30, 31) which categorizes COVID-19. Owing to their high sequence similarities with human type 3 VH domains (VH3), nanobodies are known to cause little immunogenicity (29) .

For the same reason, they can be humanized with relative ease to reduce immunogenicity when needed. Therefore, nanobodies as biotherapeutics are being 60 increasingly recognized. Examples of nanobody drugs include caplacizumab (32) for the treatment of acquired thrombotic thrombocytopenic purpura, and ozoralizumab and vobarilizumab that are in the clinical trials for rheumatoid arthritis (29, 33) . Recently, several groups have independently reported neutralizing nanobodies (22, (34) (35) (36) (37) (38) (39) or single-chain VH antibodies (40) against SARS-CoV-2 with variable potencies. 65 We have recently reported several synthetic nanobodies (sybodies) which bind RBD with various affinity and neutralizing activity (35) . Affinity and neutralizing activity are very important characteristics for therapeutic antibodies, and they can be improved by a number of ways such as random mutagenesis (22, 36) and structure-70 based design. Previously, in the case of one modestly-neutralizing sybody MR17, we have determined its structure and designed a single mutant that improves its potency by over 23 folds (35) . The rational design approach, while very effective, inevitably requires high-resolution structural information which are non-trivial to obtain.

Generally applicable tools will be welcome.

Here, we report a strategy to increase sybody potency by biparatopic fusion with SR31, a sybody that binds RBD tightly with a KD of 5.6 nM. As revealed by crystal structure, SR31 engages the RBD at a conserved site that is distal to the RBM. As such, it does not neutralize SARS-CoV-2 but forms non-competing pairs with several other RBM-binders and increases their neutralization potency when conjugated. SR31 may 80 be used as a general affinity-enhancer for both detection and therapeutic applications.

A high-affinity RBD binder without neutralizing activity 85 Previously, we generated 99 sybodies from three highly diverse synthetic libraries by ribosome and phage display with in vitro selections against the SARS-CoV-2 RBD.

Most of the sybody binders showed neutralizing activity. Interestingly, about 10 sybodies bind RBD but showed no neutralizing activities (35) even at 1 M concentration.

One such sybodies, named SR31, was characterized in this study. In analytic fluorescence-detection size exclusion chromatography (FSEC), SR31 caused earlier retention of RBD (Fig. 1A) which was included at a low concentration (0.5 M), suggesting nanomolar affinity for SR31-RBD binding. This was confirmed by bio-layer 95 interferometry analysis (Fig. 1B) which showed a KD of 5.6 nM and an off-rate of 1 × 10 -3 s -1 . Consistent with its inability to neutralize SARS-CoV-2 pseudovirus, SR31 did not affect RBD-ACE2 binding (Fig. 1C) . 

To characterize the SR31-RBD interactions in detail, we purified the complex (Fig.   1D ), and obtained crystals (Fig. 1D ) that diffracted to 1.97 Å resolution ( Table 1) . The structure was solved by molecular replacement using the published RBD and sybody 110 structures (PDB IDs 6M0J and 5M13) (6, 41) as search models. The structure was refined to Rwork/Rfree of 0.182/0.207 ( Table 1) . The asymmetric unit contained one molecule each for the RBD and SR31, indicating an expected 1:1 stoichiometry. SR31 binds to the RBD sideways at a buried surface area of 1,386.3 Å 2 ( Fig. 2A) , which is significantly larger than that for the previously reported sybodies SR4 (727.4 Å 2 ) and MR17 (853.944 Å 2 ) (35) . The binding surface is near a heavily decorated glycosylation site, Asn343 ( Fig. 2A-2C) , which, although at an apparent strategic 130 position to possibly divide the accessible surfaces for immune surveillance, does not show clashes with SR31. All three CDRs participated in the interaction by providing five (CDR1), three (CDR2), and nine H-bonds (CDR3) (Fig. 2E-2G) . Peculiarly, the CDR3, which contains a cluster of hydrophobic side chains that include Met99, Val100, Phe102, Trp103, and Tyr104, inserted into a greasy pocket (Fig. 2B ) in the RBD that 135 was lined with twelve hydrophobic/aromatic residues (Fig. 2F) . Unlike salt bridges, hydrophobic interactions are more tolerant to environment such as change of pH and ionic strength. In addition, they are less specific and thus less likely to be affected by mutations. This binding mode thus makes SR31 an attractive candidate for detection purposes.

Most RBD-targeting neutralizing antibodies, including neutralizing nanobodies characterized so far (8, 13-15, 19, 20, 22-24, 26-28, 34, 35, 37) , engage the RBD at the receptor-binding motif (RBM) (Fig. 3A) , thus competing off ACE2 and preventing viral entry. Aligning the ACE2 structure to the SR31-RBD structure showed that the SR31-145 binding epitope is distant from the RBM (Fig. 3A) . Comparing the epitopes of existing monoclonal antibodies showed that the SR31 epitope partly overlaps with CR3022 (12) and the recently identified EY6A (22) (Fig. 3B, 3C ). It has been established that the binding of the bulky CR3022 and EY6A at the interface between RBD and the N- . Taken together, the structural data rationalize the high-affinity binding between SR31 and RBD, and its inability to neutralize SARS-CoV-2.

Because nanobodies are relatively easy to produce, the availability of nanobodies that recognize a wide spectrum of epitopes can be a useful toolkit to probe binding 160 mode of uncharacterized antibodies using competitive binding assays. They may also be used to select binders with new epitopes by including them as pre-formed sybody-RBD complexes during in vitro selection (and thus excluding binders at the same site). other RBD-targeting nanobodies (22, 35, 36, 39) and mAbs (13-15, 19, 20, 23, 24, 26-28) . Red, the collective epitope of RBM-binders; blue, the SR31 epitope; magenta, the collective epitope of CR3022 and EY6A; orange, the overlap between the 

Structure alignment of SR31-RBD with ACE2-RBD revealed that the two RBD structures were overall very similar with a C RMSD of 0.452 Å (Fig. 4A) .

Nevertheless, significant structural rearrangements at the binding interface were observed (Fig. 4A, 4B) . Specifically, the small -helix 364-370 (numbers mark start-end) 185 moves towards the direction of RBM by a dramatic ~8.0 Å and transforms to a short sheet (367-370) which in turn forms a parallel -sheet pair with 102-104 in the CDR3 region. In addition, nudged by the CDR1, the short helix 383-388 swings towards the RBD core by ~4.0 Å. Remarkable, the dramatic rearrangements did not cause noticeable conformational change of RBM (Fig. 4A) nor did it affect ACE2 binding (Fig. 1C) . Given that RBD is 210 a relatively small entity, and that the two surfaces are relatively close (~25 Å), this was somewhat unexpected. A probable explanation is that RBD is very rigid and hence stable. Indeed, as shown in Fig. 4C , RBD showed ultra-stability, with an apparent melting temperature of greater than 95 º C (20-min heating).

Intriguingly, the rearrangement happens at a region that is rich in disulfide bonds.

Specifically, 367-370 is tethered between the disulfide pairs Cys379-Cys432 and Cys336-Cys361, and 383-388 bridges Cys379-Cys432 and Cys-391-Cys525 (Fig. 4D) .

Thus, the three disulfide bonds segregate the two local motifs from the rest of RBD, preventing these conformational changes from propagating through the domain.

The neutral feature of SR31 so far suggests it could bind to RBD in addition to RBM binders such as MR17 and SR4 (35) . Indeed, BLI assays showed no competition 225 between SR31 and MR17 (Fig. 5A) , indicating a 'sandwich complex' where the RBD is bound with both sybodies. This non-competing feature was also observed in the case of MR6 (Fig. 5B) which has also been shown to have neutralizing activities (35) . As a further proof for the simultaneous binding, we determined the structure of the sandwich complex SR31-RBD-MR17 (Fig. 5C, Table 1 ) to 2.10 Å resolution. The sandwich 230 complex was similar to the individual MR17-and SR31-RBD complexes, with an overall C RMSD of 0.667 and 0.375 Å, respectively. Aligning the sandwich complex with the MR17-RBD structure revealed no noticeable changes at the MR17-binding surface (Fig. 5C) , reinforcing the idea that SR31-binding does not allosterically change the RBM surface nor affect RBM binders. to the two-component complex structure (RBD (green) and MR17, PDB ID 7c8w) (35) .

Although SR31 does not neutralize SARS-CoV-2 pseudovirus itself, its highaffinity may help increase the affinity of other neutralizing nanobodies through avidity 250 effect by fusion. Indeed, the biparatopic fusion SR31-MR17 displayed remarkable increase in binding affinity compared to SR31 or MR17 alone. Its KD of 0.3 nM (Fig.   6A ) was lower than MR17 (KD = 83.7 nM) (35) by 230 folds and lower than SR31 (KD = 5.6 nM) by 17 folds. Consistently, SR31-MR17 neutralized SARS-CoV-2 pseudovirus 13 times more effectively (in molarity) than MR17 alone (Fig. 6B) . 255 That SR31 can enhance potency of its fusion partner was also demonstrated in the case for MR6. At its free form, MR6 bound to RBD with a KD of 23.2 nM (Fig. 6C) , and showed modest neutralizing activity with an IC50 of 1.32 g mL -1 (77.5 nM). Fusing it to SR31 increased its affinity by over 40 folds, displaying a KD of 0.5 nM (Fig. 6D) . 260 In line with this, SR31-MR6 showed a 27-fold higher neutralization activity compared to MR6, with an IC50 of 2.7 nM (0.08 g mL -1 ) (Fig. 6E) . Interestingly, when fused to MR3, a neutralizing antibody that had higher affinity (KD = 1.0 nM) than SR31, the neutralizing activity decreased by 2 folds (Fig. 6F) . Possible reasons include steric incompatibility caused by improper link length, and allosteric effects. Such hypothesis 265 warrants future structural investigation. Binding affinity and neutralizing activity are important characteristics of therapeutic antibodies. For modestly neutralizing nanobodies, the potency can be increased in a number of ways, including random mutagenesis (22) , structure-based 280 design (35) , and fusion (35, 36, 42) . Compared with the other two approaches, the fusion technique is more rapid, less involving and does not rely on prior structural information.

Depending on whether the two fusion partners are the same, divalent nanobodies 285 can be categorized into two types: monoparatopic and biparatopic. Biparatopic fusions recognize two distinct epitopes on the same target. Therefore, they are more likely to be resistant to escape mutants because simultaneous mutations at two epitopes should occur at a much lower rate than at a single epitope.

Because of the minute size, SR31 could be used as an 'add-on' to monoclonal antibodies, scFv fragments, and other nanobodies to enhance their affinity and potency, especially for those with modest neutralizing activities. In addition, due to its small size and high stability, SR31 may be chemically modified as a vector to deliver smallmolecule inhibitors specifically targeting SARS-CoV-2.

In summary, we have structurally characterized SR31, a high-affinity nanobody against SARS-CoV-2 RBD. Although lacking neutralizing activity alone, SR31 is an attractive biparatopic partner for RBM-binders owing to its distinct epitope from RBM.

Our work presents a generally useful strategy and offers a simple and fast approach to 300 enhance potency of modestly active antibodies against SARS-CoV2. 

The authors claim no conflict of interest.

SARS-CoV-2 RBD was expressed essentially as described (35) . Briefly, a DNA 330 fragment encoding, from N-to C-terminus, residues 330-541 of SARS-CoV2 S, a Gly-Thr linker, the 3C protease site (LEVLFQGP), a Gly-Ser linker, the Avi tag (GLNDIFEAQKIEWHE), a Ser-Gly linker, and a deca-His tag were cloned into the pFastBac-based vector. Baculovirus was generated in Sf9 cells following the Invitrogen Bac-to-Bac transfection protocol. High Five insect cells were infected with P3 virus. For crystallization, SR31 or SR31-MR17 was mixed with RBD at a 1:1.5molar ratio. The mixture was then loaded onto a Superdex 200 column for gel filtration.

Fractions containing the complex were pooled and concentrated to 10 mg mL -1 .

To screen RBD binders by size exclusion chromatography (SEC) using unpurified sybodies, RBD was fluorescently labelled as follows. First the avi-tagged RBD was for 200-300 s, before moving into sybody-free buffer for dissociation. BLI signal was monitored during the whole process. Data were fitted with a 1:1 stoichiometry using the build-in software Analysis 10.0 for kinetic parameters. For competitive assay of the 390 RBD between SR31 and ACE2, the RBD-coated sensor was saturated in 200 nM of SR31, before soaked in 25 nM SR31 with or without 25 nM of ACE2. As a control, BLI assays were also carried out by soaking the RBD-coated sensor in ACE2 without SR31. For competitive RBD-binding assays for different sybodies, the assays were carried out the same manner as described above. Desired crystals were cryo-protected, harvested using a MiTeGen loop under a microscope, and flash-cooled in liquid nitrogen before diffraction.

X-ray diffraction data were collected at beamline BL19U1 (44) at Shanghai Synchrotron Radiation Facility with a 50 x 50 μm beam on a Pilatus 6M detector, with 440 oscillation of 0.5° and a wavelength of 0.97853 Å. Data were integrated using the software XDS (45) , and scaled and merged using Aimless (46) . The SR31-RBD structure was solved by molecular replacement using Phaser (47) with PDB IDs 6M0J and 5M13 (41) as the search model. The SR31-MR17-RBD structure was solved using the SR31-RBD and MR17 structure (35) as search models. The models were manually 445 adjusted as guided by the 2Fo-Fc maps in Coot (48) , and refined using Phenix (49) .

Structures were visualized using PyMol (50).

The structure factors and coordinates were deposited in the protein data bank (PDB) under accession codes 7D2Z (SR31+RBD) and 7D30 (SR31-MR17+RBD). 

A novel coronavirus outbreak of global health concern

Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation

Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein

Cell entry mechanisms of SARS-CoV-2

SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor

Structural basis of receptor recognition by SARS-CoV-2

Structural and Functional Basis of SARS-CoV-2 Entry by Using Human ACE2

Structural basis for the recognition of SARS-CoV-2 by full-length human ACE2

Conformational dynamics of SARS-CoV-2 trimeric spike glycoprotein in complex with receptor ACE2 revealed by cryo-EM. bioRxiv

Structural basis for the neutralization of SARS-CoV-2 by an antibody from a convalescent patient

A highly conserved cryptic epitope in the receptor binding domains of SARS-480

A noncompeting pair of human neutralizing antibodies block COVID-19 virus binding to its receptor ACE2

A human neutralizing antibody targets the receptor-binding site of SARS-CoV-2

Isolation of potent SARS-CoV-2 neutralizing antibodies and protection from disease in a small animal model

Convergent antibody responses to SARS-CoV-2 in convalescent individuals

The receptor-binding domain of the viral spike protein is an 490 immunodominant and highly specific target of antibodies in SARS-CoV-2 patients

Cross-neutralization of SARS-CoV-2 by a human monoclonal SARS-CoV antibody

Potent neutralizing antibodies against multiple epitopes on SARS-CoV-2 spike. microbe

Neutralizing nanobodies bind SARS-CoV-2 spike RBD and block interaction with ACE2

Studies in humanized mice and convalescent humans yield a SARS-CoV-2 antibody cocktail

Potent neutralizing antibodies against SARS-CoV-2 identified by high-throughput single-cell sequencing of convalescent patients' B cells

A potent neutralizing human antibody reveals the N-terminal domain of the Spike protein of SARS-CoV-2 as a site of vulnerability. bioRxiv

Potent neutralizing antibodies from COVID-19 patients define multiple targets of vulnerability

Structures of Human Antibodies Bound to SARS-CoV-2 Spike Reveal 510

Structural basis for potent neutralization of SARS-CoV-2 and role of antibody affinity maturation. bioRxiv : the preprint server for biology

The Therapeutic Potential of Nanobodies

Delivery of ALX-0171 by inhalation greatly reduces respiratory syncytial virus disease in newborn lambs

Nanobodies® as inhaled biotherapeutics for lung diseases

Caplacizumab Treatment for Acquired Thrombotic Thrombocytopenic Purpura

Emerging therapies in rheumatoid arthritis: focus on monoclonal antibodies. F1000Res 8, F1000 Faculty Rev-1549

Structural Basis for Potent Neutralization of Betacoronaviruses by Single-525

Potent synthetic nanobodies against SARS-CoV-2 and molecular basis for neutralization. bioRxiv

An ultra-high affinity synthetic nanobody blocks SARS-CoV-2 infection by locking Spike into an inactive conformation. bioRxiv

An alpaca nanobody neutralizes SARS-CoV-2 by blocking receptor interaction

Synthetic nanobodies targeting the SARS-CoV-2 receptor-binding domain. bioRxiv

Selection, biophysical and structural analysis of synthetic nanobodies that 535 effectively neutralize SARS-CoV-2. bioRxiv

Identification of Human Single-Domain Antibodies against SARS-CoV-2

Synthetic single domain antibodies for the conformational trapping of membrane proteins

Generation of synthetic nanobodies against delicate proteins

The protein complex crystallography beamline (BL19U1) at the Shanghai 545

Synchrotron Radiation Facility

How good are my data and what is the resolution?

Phaser crystallographic software

Features and development of Coot

PHENIX: a comprehensive Python-based system for macromolecular structure solution