key: cord-0759074-8t04h6yl
authors: Chaouat, Abigael; Achdout, Hagit; Kol, Inbal; Berhani, Orit; Roi, Gil; Vitner, Einat B.; Melamed, Sharon; Politi, Boaz; Zahavy, Eran; Brizic, Ilija; Rovis, Tihana Lenac; Alfi, Or; Wolf, Dana; Jonjic, Stipan; Israely, Tomer; Mandelboim, Ofer
title: SARS-CoV-2 receptor binding domain fusion protein efficiently neutralizes virus infection
date: 2021-04-19
journal: bioRxiv
DOI: 10.1101/2021.04.18.440302
sha: f9402da1a1d6314cbcb63c2281fe77855d641637
doc_id: 759074
cord_uid: 8t04h6yl

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is responsible for the COVID-19 pandemic, causing health and economic problems. Currently, as dangerous mutations emerge there is an increased demand for specific treatments for SARS-CoV-2 infected patients. The spike glycoprotein on the virus membrane binds to the angiotensin converting enzyme 2 (ACE2) receptor on host cells through its receptor binding domain (RBD) to mediate virus entry. Thus, blocking this interaction may inhibit viral entry and consequently stop infection. Here, we generated fusion proteins composed of the extracellular portions of ACE2 and RBD fused to the Fc portion of human IgG1 (ACE2-Ig and RBD-Ig, respectively). We demonstrate that ACE2-Ig is enzymatically active and that it can be recognized by the SARS-CoV-2 RBD, independently of its enzymatic activity. We further show that RBD-Ig efficiently inhibits in vitro and in vivo SARS-CoV-2 infection, better than ACE2-Ig. Mechanistically we show that anti-spike antibodies generation, ACE2 enzymatic activity and ACE2 surface expression were not affected by RBD-Ig. Finally, we show that RBD-Ig is more efficient than ACE2-Ig at neutralizing high virus concentration infection. We thus propose that RBD-Ig physically blocks virus infection by binding to ACE2 and that RBD-Ig should be used for the treatment of SARS-CoV-2-infected patients. Author Summary SARS-CoV-2 infection caused serious socio-economic and health problems around the globe. As dangerous mutations emerge, there is an increased demand for specific treatments for SARS-CoV-2 infected patients. SARS-CoV-2 infection starts via binding of SARS-CoV-2 spike protein receptor binding domain (RBD) to its receptor, ACE2, on host cells. To intercept this binding, we generated Ig-fusion proteins. ACE2-Ig was generated to possibly block RBD by binding to it and RBD-Ig to block ACE2. We indeed showed that the fusion proteins bind to their respective target. We found that it is more efficient to inhibit SARS-CoV-2 infection by blocking ACE2 receptor with RBD-Ig. We also showed that RBD-Ig does not interfere with ACE2 activity or surface expression. Importantly, as our treatment does not target the virus directly, it may be efficient against any emerging variant. We propose here that RBD-Ig physically blocks virus infection by binding to ACE2 and thus it may be used for the treatment of SARS-CoV-2-infected patients.

assess inhibition by ACE2-Ig we initially incubated increasing concentration of the 145 fusion protein with 300 PFU/ml of SARS-CoV-2 for 1 hour at 37°C, and then infected 146 Vero E6 cells. Conversely, to test for inhibition by RBD-Ig, we had to first incubate the 147 fusion protein with Vero E6 cells for 1 hour at 37°C, and then infect with 300 PFU/ml of 148 SARS-CoV-2. Both strategies required a 48-hour incubation period to allow for plaque the quality of the antibodies, as we suspected that RBD-Ig treated mice will generate 189 more neutralizing antibodies since it was shown that anti-RBD antibodies have 190 neutralization effect [27] . To test this hypothesis, we used the 293T-Spike cells and 191 stained them with ACE2-Ig in the presence or absence of sera obtained from all mice 192 groups. Since naïve mice did not generate anti-spike antibodies ( Figure 4A ), their sera, as 193 expected, did not contained neutralizing antibodies. Indeed, similar ACE2-Ig binding was 194 observed with and without blocking ( Figure 4B ). In all the infected mice, a comparable 195 level of blocking was seen with sera regardless of the treatment administered, as assessed 196 by reduced ACE2-Ig staining ( Figure 4B ). From these results we concluded that all the 197 antibodies that were generated were a result of SARS-CoV-2 infection rather than our 198 treatment. To further investigate why is RBD-Ig better than ACE2-Ig at inhibiting SARS-CoV-2 202 infection we first generated a specific monoclonal antibody against ACE2 using ACE2-Ig 203 as an antigen. This step was essential since the commercial antibodies (#ABIN1169449 204 and #MA5-32307) we tested did not recognize ACE2 effectively. As can be seen in 205 Figure 5A , our generated antibody (anti-ACE2 01) is specific to ACE2, as it binds only to 206 the 293T-ACE2 cells. To test if the antibody blocks the interaction with SARS-CoV-2 207 RBD, we incubated the antibody with 293T-ACE2 cells for 1 hour and then stained the 208 cells with RBD-Ig. The anti-ACE2 01 antibody has no blocking property as its presence 209 did not interfere with the binding of RBD-Ig to the ACE2 ( Figure 5B ). 210 We then analyzed whether the expression of ACE2 is altered, at various time points, 211 following SARS-CoV-2 infection. We infected 293T-ACE2 cells with a 0.5 MOI and 212 compared between ACE2 surface expression on infected cells to uninfected cells using 293T-ACE2 cells will lead to reduced ACE2 surface expression as we hypothesized that 220 this might be the reason why RBD-Ig is more efficient than ACE2-Ig at neutralizing 221 infection. We incubated RBD-Ig or Control-Ig with 293T-ACE2 cells for 1, 2, 6 and 24 222 hours. Following incubation cells were harvested and ACE2 surface expression was 223 assessed by flow cytometry using our generated antibody, anti-ACE2 01. As can be seen 224 in Figure 5D , ACE2 surface levels were only slightly reduced following RBD-Ig binding, 225 suggesting that this is not the reason why RBD-Ig is superior to ACE2-Ig. 226 Next, we examined whether ACE2 activity will be altered following interaction with 227 RBD-Ig, as we thought that maybe the activity of ACE2 might affect somehow the 228 infection. We incubated 0.1 or 1 ug of RBD-Ig or Control-Ig with recombinant human 229 ACE2 or with 293T-ACE2 cells lysate. ACE2 activity was not affected by RBD-Ig when 230 incubated with human ACE2 or with a lysate containing ACE2 ( Figure 5E ). These combined results suggest that treatment with RBD-Ig inhibits infection without affecting 232 ACE2 activity and surface levels expression.

Our last assumption was that RBD-Ig inhibits infection by physically blocking ACE2. 234 We further hypothesized that RBD-Ig is more efficient than ACE2-Ig because RBD-Ig . 273 We showed that ACE2-Ig inhibits in vitro SARS-CoV-2 infection as it has been 274 previously shown [42] and that RBD-Ig inhibits infection significantly more than ACE2-

Ig. Furthermore, we show that treatment with RBD-Ig using SARS-CoV-2 K18-hACE2 276 infected mice led to decrease in disease severity as assessed by reduced body weight and increased mice survival. Importantly, 50% of the RBD-Ig treated mice survived although 278 active infection occurred, while ACE2-Ig injection had no effect. We think that the 279 reason behind the low efficiency of ACE2-Ig in vivo is due to the low concentrations of 280 fusion protein we administered which was 75ug/mouse, injected intraperitoneally. . 283 We demonstrated that the superiority of RBD-Ig was not due to quantitative or qualitative 284 changes in the antibody response, we thus hypothesized that it may be due to RBD-Ig 285 effect on its target protein ACE2. To check this, we first wanted to assess whether perhaps because there was no effective commercial antibody available against ACE2. 291 We also assessed ACE2 surface expression following RBD-Ig incubation and saw that 292 ACE2 expression did not change drastically. Another important check was of ACE2 293 enzymatic activity following binding to SARS-CoV-2 RBD as it was reported to enhance 294 ACE2 activity [47] . In contrast, we report here that ACE2 activity was not affected 295 following incubation with RBD-Ig. The reason for this discrepancy is not understood. 296 We next hypothesized that RBD-Ig blocks infection by physically interacting with ACE2. 297 We further thought that RBD-Ig is more efficient than ACE2-Ig since RBD-Ig, binds to 298 the constantly expressed ACE2 on the target cells, while ACE2-Ig interacts with the spike protein found on a replicating virus. Indeed, we showed that RBD-Ig can neutralizes in 300 vitro SARS-CoV-2 even at high virus titers, while ACE2-Ig cannot.

To summarize we suggest that RBD-Ig inhibit SARS-CoV-2 infection by physically 302 blocking ACE2. Thus, RBD-Ig is particularly advantageous as a treatment for SARS- 

Vero E6 cells (CRL-1586, ATCC) were seeded in 12-well plates (5x10 5 cells/well) and 415 grown overnight in Penicillin-Streptomycin-Neomycin (P/S/N, BI) containing medium.

The following day, ACE2-Ig and Control-Ig were either diluted to 50µg/ml-0.048µg/ml 

Angiotensin-Converting Enzyme 2 Metabolizes and Partially Inactivates Pyr-594

Apelin-13 and Apelin-17: Physiological Effects in the Cardiovascular System

Angiotensin1-9 599 antagonises pro-hypertrophic signalling in cardiomyocytes via the angiotensin type 600 2 receptor

A 603 novel angiotensin-converting enzyme-related carboxypeptidase (ACE2) converts 604

angiotensin I to angiotensin 1-9

An Evidence-Based Narrative Review of Treatment Options

Therapeutic Fc-fusion proteins and peptides as successful 611 alternatives to antibodies

Transmission 672 of SARS-CoV-2 Lineage B.1.1.7 in England: Insights from linking 673 epidemiological and genetic data. medRxiv

The 677 E484K mutation in the SARS-CoV-2 spike protein reduces but does not abolish 678 neutralizing activity of human convalescent and post-vaccination sera

SARS-CoV-2 501Y.V2 escapes neutralization by South African COVID-19 684 donor plasma

Bamlanivimab as Monotherapy or in Combination With Etesevimab on Viral Load 688 in Patients With Mild to Moderate COVID-19