key: cord-0712247-g887losk
authors: Weinstein, Jules B.; Bates, Timothy A.; Leier, Hans C.; McBride, Savannah K.; Barklis, Eric; Tafesse, Fikadu G.
title: A potent alpaca-derived nanobody that neutralizes SARS-CoV-2 variants
date: 2022-02-22
journal: iScience
DOI: 10.1016/j.isci.2022.103960
sha: bbe358bba7905938f12ef40653ef8ffb5a0d1313
doc_id: 712247
cord_uid: g887losk

The spike glycoprotein of SARS-CoV-2 engages with human angiotensin-converting enzyme 2 (ACE2) to facilitate infection. Here, we describe an alpaca-derived heavy chain antibody fragment (VHH), saRBD-1, that disrupts this interaction by competitively binding to the spike protein receptor-binding domain. We further generated an engineered bivalent nanobody construct engineered by a flexible linker, and a dimeric Fc conjugated nanobody construct. Both multivalent nanobodies blocked infection at picomolar concentrations and demonstrated no loss of potency against emerging variants of concern including Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), Epsilon (B.1.427/429), and Delta (B.1.617.2). saRBD-1 tolerates elevated temperature, freeze-drying, and nebulization, making it an excellent candidate for further development into a therapeutic approach for COVID-19.

The COVID-19 pandemic, caused by severe acute respiratory syndrome coronavirus-2 (SARS-24

CoV-2), is an ongoing global health crisis with over 230 million cases, 4.8 million deaths world-25 wide as of October 2021 (Dong et al., 2020) . While several effective vaccines have been 26 developed, concern about potential future surges of infections remain, due to the proliferation 27 and spread of multiple variant strains, combined with waning protection from vaccination (Levin 28 et al., 2021; Shrotri et al., 2021) . It is anticipated that additional variants will continue to emerge, 29 and the slow pace of global vaccination creates greater opportunity for emergence and spread 30 of vaccine resistant variants (Luo et al., 2021) . 31 32 SARS-CoV-2 is an enveloped, positive-sense, single-stranded RNA virus, and a member of the 33 Coronaviridae family, so named for the crown-like protrusions visible on their outer membranes 34 in EM micrographs (Huang et al., 2020) . Four structural proteins are encoded by SARS-CoV-2: 35 spike (S), envelope, membrane, and nucleocapsid . Homotrimers of the S 36 glycoprotein form the characteristic crown-like protrusions on the virion surface, where it 37 facilitates entry into cells through its interaction with the cell surface protein angiotensin-38 converting enzyme 2 (ACE2) (Hoffmann et al., 2020) . Each monomer of S is composed of two 39

Neutralizing antibodies have been shown to be protective against COVID-19 disease (Khoury et showed no detectable binding, demonstrating that saRBD-1 binds specifically to the RBD 126 subunit of the S protein ( Figure 2B, D) . Due to the promising initial binding characteristics of 127 saRBD-1, we next investigated the binding kinetics in greater detail using bio-layer 128 interferometry (BLI), which measures the effective mass change at the surface of a sensor tip. 129

As expected, the S2 protein control yielded no binding ( Figure 1C, D) . However, BLI tips loaded 130 with RBD measured a dissociation constant (KD) of 750 pM for saRBD1, while tips coated with 131 S1 yielded a KD 1880 pM. S trimer loaded tips showed the strongest binding with a KD of 674 132 pM, consistent with our ELISA results. These KD's are lower than the previously reported 15 nM 133 KD of the RBD-ACE2 interaction, suggesting that saRBD-1 binds SARS-CoV-2 with at least an 134 order of magnitude greater affinity than ACE2 (Glasgow et al., 2020) . 135

To more thoroughly examine this, we performed BLI-based competition assays to determine if 137 saRBD-1 is able to block the RBD-ACE2 interaction ( Figure 2E ). In this assay, BLI tips were first 138 loaded with RBD followed by varying concentrations of saRBD-1 VHH to block the RBD binding 139 sites before finally transferring to a solution with a fixed concentration of ACE2. We found that 140 saRBD-1 bound competitively with ACE2, and that a concentration of 6 nM of saRBD-1 was 141 sufficient to block 50% of ACE2 binding. These results indicated that saRBD-1 binds specifically 142 to the RBD subunit of native trimeric S protein with picomolar affinity and blocks the subsequent 143 interaction of RBD with ACE2. To determine VHH inhibitory activities against live SARS-CoV-2 virus, focus forming assays 163 were performed using SARS-CoV-2 WA1/2020 strain and saRBD-1. For the assay, Vero E6 or 164 human colorectal epithelial (Caco-2) cells were infected with SARS-CoV-2, then stained with 165 anti-S alpaca polyclonal sera as a primary antibody and an HRP-conjugated secondary 166 antibody, facilitating visualization of SARS-CoV-2 infected cells ( Figure 3C , D). The 50% focus 167 reduction neutralization titer (FRNT50) was found to be 5.82 nM for Vero E6, and 7.4 nM for 168 Caco2 ( Figure 3E ). In comparison, the non-neutralizing VHH52 failed to decrease foci. Thus, it 169 is evident that monovalent saRBD-1 is a potent neutralizer of live SAR-COV-2 in vitro, even at 170 low nanomolar concentrations. To test this with saRBD-1, we  178   utilized a mammalian vector to express saRBD-1 conjugated to human IgG Fc with a short  179 hinge (Hanke et al., 2020; Tiller et al., 2008) . The resulting chimeric protein is secreted as a 180 dimer due to disulfide bridging of two Fc regions, and thus acts as a partially humanized heavy-181 chain only antibody ( Figure 4A ). This approach allows for improved binding due to avidity effects 182 and greater steric blockage of the ACE2 binding site of the S protein. Simultaneously, we 183 produced a bivalent construct of saRBD-1 (BI-saRBD-1) attached by a flexible (GGGGS)4 linker 184 (Shan et al., 1999; Wrapp et al., 2020a) . To determine binding kinetics of the saRBD-1 Fc-dimer 185 (Fc-saRBD-1) to RBD, we utilized ELISA and BLI ( Figure 4B -C, Figure S2 ). The EC50 of Fc-186 saRBD-1 as measured by ELISA was 392 pM, a 50% stronger affinity as compared to 187 monovalent saRBD-1. The KD of Fc-saRBD-1 as measured by BLI was 302 pM, primarily driven 188 by a 3-fold reduction in the KOFF compared to monovalent saRBD-1. Using our pseudovirus 189 neutralization assay, the neutralization ability of the Fc-saRBD-1 dimer improved to an IC50 of 190 100 pM, over a 40-fold improvement compared to monomeric saRBD-1 ( Figure 4D We evaluated the stability of saRBD-1 by subjecting it to some of the conditions that are likely to 205 be encountered during production, transport, and delivery of protein-based therapeutics we 206 evaluated the stability of saRBD-1 in elevated temperature, lyophilization, and nebulization. We 207 treated VHH to each condition, then measured of protein loss, binding kinetics, and neutralizing 208 ability of the treated VHH aliquots. Aliquots of saRBD-1 were incubated for 1 hour at 50°C then 209 centrifuged to remove aggregates before measurement of protein loss by OD280, which showed 210 a 19% reduction. The treated aliquots were then checked by BLI on RBD ( Figure 5G , I), which 211 showed minimal loss of activity concomitant with the reduction in measured protein 212 concentration. Similar measurements were performed using lyophilized (29% protein loss) and 213 nebulized (77% protein loss) samples. Nebulization is known to be a harsh process, particularly 214 when performed in unmodified PBS solution with a jet nebulizer, and our numbers mirror 215 previous reports of 4-fold loss of activity after nebulization with an ultrasonic nebulizer (Schoof 216 et al., 2020) . In total, we found that the KD was 938 pM for heat treatment, 936 pM for 217 lyophilized, and 3.65 nM for aerosolized, amounting to 1.25-fold, 1.25-fold, and 4.8-fold 218 increases respectively, which align with our protein loss determinations. 219 220 To assay effects of these treatments on neutralizing activity, we carried out focus forming 221 assays in VeroE6 cells utilizing the heat treated, lyophilized, and nebulized saRBD-1 samples 222 ( Figure 3H , J). We found that 50°C treated, lyophilized, and nebulized saRBD-1 yielded 223

FRNT50s of 3.00 nM, 3.74 nM, and 9.01 nM respectively. In comparison, untreated saRBD-1 224 yielded a FRNT50 of 5.82 nM. Therefore, only nebulization reduced saRBD-1 neutralizing 225 capability, with a 1.56-fold reduction. Overall saRBD-1 appears functionally stable, and it 226 maintains nanomolar neutralization activity towards RBD even after destabilizing treatments. 227 228 229 J o u r n a l P r e -p r o o f

Because of the prevalence of SARS-CoV-2 variant strains of concern (VOCs) significantly 231 divergent from the base strain ( Figure 6A ), we sought to test saRBD-1's affinity for mutated 232 RBD-N501Y and neutralizing abilities against clinical VOC isolates We generated a variant RBD 233 to test saRBD-1-RBD interactions. Using site directed mutagenesis, we created a spike and 234 RBD variant that contained the N501Y mutation found in several of the circulating VOCs ( Figure  235 6B). Using BLI, we found binding of saRBD-1 to RBD-N501Y was similar to WT saRBD-1 236 ( Figure 6C interferometry (BLI) competitive binding assay. In this assay, SARS-CoV-2 spike RBD protein 270 was attached to a sensor and first exposed to saRBD-1, which bound strongly during the first 271 300 seconds. In the subsequent step, the sensors were transferred to solutions containing the 272 representative monoclonal antibodies. SaRBD-1 successfully blocked B38 from binding, 273 indicating that they likely bind to overlapping epitopes ( Figure 7B ). The class 2 & 3 antibodies 274 were not affected by saRBD-1. A control experiment confirmed that B38 binds successfully 275 when saRBD-1 was absent ( Figure 7C ). These results were recapitulated with a dimeric Fc-276 saRBD-1 construct (Figure7D) . Hence, saRBD-1 is most likely a class 1 binder. An unlikely 277 alternative is that saRBD-1 binds a distal site non-competitive with the class 3 antibody, but 278 forces RBD into a down conformation unsuitable for B38 binding. therapeutics and vaccines should fulfill both the following conditions: 1) be affordable to 296 produce, transport and store. 2) provide highly effective long-term protection against circulating 297

VOCs. Our saRBD-1 VHH is an ideal match due to its cheap manufacture of bacterial 298 purification, thermostability, and efficacy at VOC neutralization. 299

The ability of saRBD-1 to potently neutralize SARS-CoV-2 is critical to its potential. Antiviral Interestingly, saRBD-1 neutralizes VOCs containing K417, E484, and N501 mutations that 360 typically affect class 1 and 2 antibodies, suggesting its epitope identity or mechanism of 361 neutralization may be atypical for class 1 neutralizing antibodies. The recently published VHH 362

Fu2 is an example of an atypical mechanism of SARS-CoV-2 neutralization (Hanke et al., 363 2022) . Fu2 binds as a class 1 antibody to block ACE2 biding to RBD, yet simultaneously 364 induces dimerization in full-length spike to further disrupt ACE2 interactions. Fu2 was also found 365 to neutralizes the Beta variant without significant loss of potency. Thus, a VHH may have 366 unexpected levels of utility outside of those predicted by epitope class, which can aid in binding 367 mutated RBD variants. 368

We found that saRBD-1 binds competitively with human ACE2 for SARS-CoV-2 spike RBD, and 370 that pre-incubation of RBD with saRBD1 blocks ACE2 binding, a necessary step to infection. 371

Low nanomolar concentrations of monovalent saRBD-1 successfully neutralize clinical isolates 372 of the Alpha, Beta, Gamma, Epsilon, and Delta VOC as a likely class 1 antibody. Both the Bi-373 saRBD-1 and Fc-saRBD-1 demonstrate improved binding, and they successfully neutralize the 374 variants at picomolar concentrations with no discernable loss of potency. Due to its high 375 neutralizing efficacy, saRBD-1, alone or in combination with other ultra-potent VHHs, is an 376 excellent candidate for development into a therapeutic to manage severe COVID-19. 377

Although saRBD-1 effectively neutralizes all the SARS-CoV-2 variants of concern that we 379 tested, it remains to be determined if it neutralizes future emerging variants. We demonstrated 380 that saRBD-1 competitively binds SARS-CoV-2 RBD with a class 1 neutralizing antibody B38, 381

indicating the potential of a class 1 epitope for saRBD-1. However, this study does not present 382 structural data for the saRBD-1-RBD complex. As such, we do not have confirmation for the 383 J o u r n a l P r e -p r o o f mechanism of SARS-CoV-2 neutralization. 385

Acknowledgments 386 BLI data were generated on an Octet Red 384, which is made available and supported by 387 OHSU Proteomics Shared Resource facility and equipment grant number S10OD023413. We 388 also thank the OHSU Flow Cytometry Shared Resource, and OHSU Advanced Light 389

Microscopy Core for the use of their software, equipment, and expertise. This of saRBD-1 on plates coated with SARS-CoV-2 RBD, S1, S2, and full-length S trimer where 420 saRBD-1 is seen to bind RBD, S1, and full-length S trimer, but not S2. Curves show the 421 average of 3 replicate experiments. C) Representative BLI curves of saRBD-1 binding kinetics 422 experiments on SARS-CoV-2 S RBD, S1, S2, and full-length trimer where saRBD-1 is seen to 423 bind RBD, S1, and full-length S trimer, but not S2. Biotinylated spike constructs were pre-bound 424 to streptavidin biosensor tips, after which association and dissociation steps were carried out in The neutralization protocol was based on previously reported neutralization methods utilizing 718 SARS-CoV-2 S pseudotyped lentivirus (Crawford et al., 2020) . 293T-ACE2 cells were seeded 719 poly-lysine treated 96-well plates at a density of 10,000 cells per well. Cells were allowed to 720 grow overnight at 37°C. LzGreen SARS-COV-2 S pseudotyped lentiviruses were mixed with 721 saRBD-1, or VHH52 control antibody. Immunized alpaca serum was used as positive 722 neutralization control, while virus alone was used as negative control. Dilutions of antibodies 723 ranged from 177 nM to 170 pm for saRBD-1 and 26.3nM and 25 pM for Fc-saRBD-1, and 6.57 724 nM to 4 pM Bi-saRBD-1. Virus-antibody mixture was incubated at 37C for 1 hour after which 725 polybrene was added up to 5 μg/ml and the mixture was added to 293T-ACE2 cells. Cells were 726 incubated with neutralized virus for 44 hours before imaging. Cells were fixed with 4% PFA for 1 727 hour at RT. Fixed cells were washed with PBS 2x, then incubated with 10 μg/ml DAPI for 10 728 minutes at RT, imaged with BZ-X700 all-in-one fluorescent microscope (Keyence). Estimated 729 area of DAPI and GFP fluorescent pixels were calculated with built in BZ-X software (Keyence). 730 731 Focus forming assay (FFA) 732

The FFAs was performed as previously described (Case et al., 2020) . In brief, Vero E6 cells 733 were plated at 20,000 cells/well or Caco-2 cells were plated at 24,000 cells/well in 96-well plates 734 and incubated overnight. Titrated SARS-CoV-2 stocks were diluted to 3,333 ffu/mL. To 20 μL of 735 virus, 20 μL of antibody dilutions were added: saRBD-1, VHH52, or Fc-saRBD-1 were used at 736 8×4-fold serial dilutions ranging from 6.25 μg/mL to 381 pg/mL for saRBD-1, 1.25 μg/mL to 76.2 737 pg/mL for Fc-saRBD-1, and 420 μg/mL to 25.6 pg/mL All virus and antibody dilutions were 738 For ELISA and neutralization data, EC50 and IC50 values were calculated using python software 750 pipeline based on input data. Curves were fit to each data set using the same pipeline. 751

For ELISA data, EC50 were calculated from OD450 nm signal relative to maximal signal for a 752 given pattern. Background was subtracted, then each was normalized to the maximum value for 753 that antigen. The S2 domain data was analyzed differently, as it was comparable to 754 background, background absorbance was first subtracted before normalization to maximum 755 value. L 1 8 F T 2 0 N K 4 1 7 T N 5 0 1 Y H 6 5 5 Y D 6 1 4 G E 4 8 4 K D 1 3 8 Y R 1 9 0 S P 2 6 S V 1 1 7 6 F T 1 0 2 7 T 1 9 R L 4 5 2 R P 6 8 1 R D 6 1 4 G T 4 7 8 K d e l1 5 6 / 1 5 7 R 1 5 8 G G 1 4 2 D D 9 5 0 N S 1 3 I W 1 5 2 C L 4 5 2 R D 6 1 4 G S 1 3 I L 4 5 2 R D 6 1 4 G 

A serological 776 assay to detect SARS-CoV-2 seroconversion in humans

Efficacy and Safety of the mRNA

CoV-2 Vaccine

SARS-CoV-2 782 neutralizing antibody structures inform therapeutic strategies

Neutralization of SARS-CoV-2 variants 785 by convalescent and BNT162b2 vaccinated serum

Age-Dependent Neutralization of SARS-CoV-2 and P.1 Variant by Vaccine 788 Immune Serum Samples

Cross-reactivity of SARS-CoV structural protein antibodies against SARS-CoV-2

Antibody Response and Variant Cross-Neutralization After SARS-CoV-2 794 Breakthrough Infection

Design, Expression

Characterization of a YFP-Tagged 2019-nCoV Spike Receptor-Binding Domain 797

Humoral immune responses and neutralizing antibodies against SARS-CoV-2; implications in 800 pathogenesis and protective immunity

Neutralizing Antibody and Soluble ACE2 Inhibition of a 803

Replication-Competent VSV-SARS-CoV-2 and a Clinical Isolate of SARS-CoV-2

Rapid and Direct VHH and Target Identification by Staphylococcal Surface 806 Display Libraries

Resistance of SARS-CoV-2 variants to neutralization by 809 monoclonal and serum-derived polyclonal antibodies

Protocol and reagents for pseudotyping 812 lentiviral particles with SARS-CoV-2 Spike protein for neutralization assays (Microbiology)

Addressing the Cold Reality of mRNA Vaccine Stability

Early Evidence of the Effect of 817 SARS-CoV-2 Vaccine at One Medical Center

Transmission, infectivity, 821 and neutralization of a spike L452R SARS-CoV-2 variant

An interactive web-based dashboard to track COVID-823 19 in real time

Faster Isolation of PBMC Using Ficoll-Paque® Plus in the Eppendorf 825 Multipurpose Benchtop Centrifuges 5920 R and 5910 Ri

827 Implications of antibody-dependent enhancement of infection for SARS-CoV-2 828 countermeasures

Protein and Antibody Engineering by Phage Display

Engineered ACE2 receptor traps potently neutralize SARS-833 CoV-2

A SARS-CoV-2-Human Protein-836

Protein Interaction Map Reveals Drug Targets and Potential Drug-Repurposing (Systems 837 Biology)

Complete Mapping of Mutations to the 840 SARS-CoV-2 Spike Receptor-Binding Domain that Escape Antibody Recognition

Fusion of the 843 mouse IgG1 Fc domain to the VHH fragment (ARP1) enhances protection in a mouse model of 844 rotavirus

Neutralization of SARS-CoV-2 847 by highly potent, hyperthermostable, and mutation-tolerant nanobodies

An alpaca 850 nanobody neutralizes SARS-CoV-2 by blocking receptor interaction

A bispecific monomeric nanobody induces spike 854 trimer dimers and neutralizes SARS-CoV-2 in vivo

SARS-CoV-2 Cell Entry 857 Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

SARS-CoV-2 variants B.1.351 and P.1 861 escape from neutralizing antibodies

Tracking Coronavirus Vaccinations Around the World. The New York Times

Effectiveness of 864 Convalescent Plasma Therapy for COVID-19 Patients in Hunan

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. The 868 Lancet

Exploiting Nanobodies' Singular Traits

Neutralizing Antibodies against SARS-CoV-2 and Other 872 Human Coronaviruses

The neutralizing 875 antibody, LY-CoV555, protects against SARS-CoV-2 infection in nonhuman primates

Logistical challenges for

CoV-2 vaccine and a call to research institutions, developers and manufacturers

Viridot: An automated virus 882 plaque (immunofocus) counter for the measurement of serological neutralizing responses with 883 application to dengue virus

Neutralizing antibody levels are highly 886 predictive of immune protection from symptomatic SARS-CoV-2 infection

A therapeutic neutralizing antibody targeting receptor binding domain of 889 SARS-CoV-2 spike protein

Structure-guided multivalent 892 nanobodies block SARS-CoV-2 infection and suppress mutational escape

Tracking Changes in SARS-CoV-2 Spike: 895 Evidence that D614G Increases Infectivity of the COVID-19 Virus

Current status of therapeutic monoclonal 897 antibodies against SARS-CoV-2

Universal protection against 900 influenza infection by a multidomain antibody to influenza hemagglutinin

Waning Immune Humoral Response to BNT162b2 Covid-904 19 Vaccine over 6 Months

The basis of a more contagious 501Y.V1 variant of SARS-CoV-2

BNT162b2-elicited neutralization of B.1.617 and other SARS-CoV-2 910 variants

Striking Antibody Evasion Manifested by the Omicron Variant of SARS-913

CoV-2. Nature 1-8

SARS-CoV-2 biology and 915 variants: anticipation of viral evolution and what needs to be done

Alpaca (Lama pacos) 918 as a convenient source of recombinant camelid heavy chain antibodies (VHHs)

Highly synergistic combinations of nanobodies 922 that target SARS-CoV-2 and are resistant to escape

A global database of COVID-19 vaccinations

Cross-neutralization of SARS-CoV-2 by a human 927 monoclonal SARS-CoV antibody

Sensitivity of infectious SARS-CoV

B.1.1.7 and B.1.351 variants to neutralizing antibodies

Nanobody cocktails potently neutralize SARS-CoV-2 933 D614G N501Y variant and protect mice

Fusion of hIgG1-Fc to 936 111In-anti-amyloid single domain antibody fragment VHH-pa2H prolongs blood residential time 937 in APP/PS1 mice but does not increase brain uptake

Single Domain Antibodies Derived from Dromedary Lymph Node and Peripheral Blood 941

Lymphocytes Sensing Conformational Variants of Prostate-specific Antigen

Nanobody: outstanding features for 944 diagnostic and therapeutic applications

Phenotypic Lentivirus Screens to Identify Antiviral Single Domain 946

Influenza Virus: Methods and Protocols

An ultrapotent synthetic 950 nanobody neutralizes SARS-CoV-2 by stabilizing inactive Spike

Characterization of 952 scFv-Ig Constructs Generated from the Anti-CD20 mAb 1F5 Using Linker Peptides of Varying 953 Lengths

Neutralization of SARS-CoV-2 Variants B.1.429 and B.1.351

Spike-antibody waning after second dose of 959 BNT162b2 or ChAdOx1

Efficient generation of monoclonal antibodies from single human B cells by single cell RT-PCR 962 and expression vector cloning

Biparatopic nanobodies protect mice from lethal 965 challenge with SARS-CoV-2 variants of concern

REGN-COV2, a Neutralizing Antibody Cocktail, in Outpatients 968 with Covid-19

Structural Basis for Potent 971 Neutralization of Betacoronaviruses by Single-Domain Camelid Antibodies

Structural Basis for Potent 975 Neutralization of Betacoronaviruses by Single-domain Camelid Antibodies

A noncompeting pair of human neutralizing antibodies block COVID-19 virus binding to 979 its receptor ACE2

Versatile and multivalent nanobodies efficiently neutralize SARS-982 CoV-2

Nanobodies from camelid mice and llamas neutralize 985 SARS-CoV-2 variants

A highly conserved cryptic epitope in the receptor binding domains of SARS-CoV-2 and 988 SARS-CoV

The efficacy assessment of convalescent plasma therapy for COVID-19 patients: a 991 multi-center case series

SARS-CoV-2 Omicron strain exhibits potent capabilities for immune evasion and viral 994 entrance

Engineered Multivalent Nanobodies Potently and Broadly Neutralize SARS-CoV-2 Variants

CoV-2 variants effectively neutralized by saRBD-1 VHH with picomolar affinity.  saRBD-1 neutralization increases when expressed as a bivalent or Fc construct.  saRBD-1 binds SARS-CoV-2 RBD as a likely class 1 neutralizing antibody.  saRBD-1 retains binding

J o u r n a l P r e -p r o o f Key resource table REAGENT