A Neutralization Assay Based on Pseudo-Typed Lentivirus with SARS CoV-2 Spike Protein in ACE2-Expressing CRFK Cells

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a highly pathogenic zoonotic virus that spreads rapidly. In this work, we improve the hitherto existing neutralization assay system to assess SARS-CoV-2 inhibitors using a pseudo-typed lentivirus coated with the SARS-CoV-2 spike protein (LpVspike +) and angiotensin-converting enzyme 2 (ACE2)-transfected cat Crandell–Rees feline kidney (CRFK) cells as the host cell line. Our method was 10-fold more sensitive compared to the typical human embryonic kidney 293T (HEK293T) cell system, and it was successfully applied to quantify the titers of convalescent antisera and monoclonal anti-spike antibodies required for pseudo virus neutralization. The 50% inhibition dilution (ID50) of two human convalescent sera, SARS-CoV-2 immunoglobulin G (IgG) and SARS-CoV-2 immunoglobulin M (IgM), which were 1:350 (±1:20) and 1:1250 (±1:350), respectively. The 50% inhibitory concentration (IC50) of the IgG, IgM and immunoglobulin A (IgA) anti-SARS-CoV-2 monoclonal antibodies (mAbs) against LpVspike(+) were 0.45 (±0.1), 0.002 (±0.001) and 0.004 (±0.001) µg mL−1, respectively. We also found that reagents typically used to enhance infection were not effective in the CFRK system. This methodology is both efficient and safe; it can be employed by researchers to evaluate neutralizing monoclonal antibodies and contribute to the discovery of new antiviral inhibitors against SARS-CoV-2.


Introduction
An outbreak of severe acute respiratory syndrome (SARS) from 2002 to 2003 in Guangdong, China, resulted in over 8000 infections and almost 800 deaths. This was caused by a coronavirus (CoV), known as SARS-associated CoV (SARS-CoV) [1]. In December 2019, a novel infectious disease caused by a related coronavirus, SARS-CoV-2, was reported in Wuhan, China. It is widely referred to as coronavirus disease 2019 (COVID-19) [2]. SARS-CoV-2 belongs to the same species as SARS-CoV and other SARS-related bat CoVs [3]. COVID-19 was declared a worldwide pandemic by the World Health Organization (WHO), and has infected over 51 million people and killed more than 1.2 million (December 2020).

Measuring LpVspike (+) Infectivity in Various ACE2-Expressing Cell Types
Although ACE2-expressing HEK293T cells are typically used with the pseudo-typed lentiviral system, the sensitivity of infectivity assays is low. With the goal of identifying a Figure 1. Generation of a pseudo-typed lentivirus coated with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Spike protein using human embryonic kidney 293T (HEK293T) cells, and illustration of the neutralization assay principle using angiotensin-converting enzyme 2 (ACE2)-expressing Crandell-Rees feline kidney (CRFK) cells. The pseudotyped lentivirus coated with SARS-CoV-2 Spike (LpVspike(+)) and harboring the 2019-nCov_pcDNA3.1(+)-P2A-eGreen Fluorescent Protein (GFP) plasmid was prepared by co-transfecting 5 × 10 5 cell mL −1 of HEK293T cells with the HIV-1 NL4-3 ∆Env ∆Vpr Luciferase Reporter Vector. Transfected HEK293T cells were incubated at 37 • C under 5% CO 2 for 2 days. The supernatant medium containing LpVspike(+) was harvested and filtered through a 0.45-µm pore size filter. To generate ACE2-expressing cell lines, cells were transfected with pcDNA3.1+/C-(K) DYK-ACE2 and incubated at 37 • C under 5% CO 2 for 1 day. The transfected cells were then able to transiently express the ACE2 receptor on their membrane surface and could hence be infected by LpVspike(+). Approximately 2.3 × 10 4 ACE2-expressing cells were infected per well in 96-well plates, and luciferase expression was measured at 48 h post-infection. This figure was created by biorender.com.
The TZM-bl cell line (also called JC.53bl-13) is a HeLa cell derivative engineered to express CD4, CCR5 and CXCR4, as well as Tat-responsive firefly luciferase reporter genes. These features make TZM-bl cells highly susceptible to HIV-1 infection and enables quantification of viral infection and neutralizing antibody titers [17]. Because TZM-bl already has an integrated copy of the luciferase reporter gene, we constructed a second pseudotyped lentivirus, named LpVspike2(+); this was essentially the same as LpVspike(+), except it did not contain the luciferase gene. ACE2-expressing TZM-bl cells infected with undiluted LpVspike2(+) produced a reading of 2 × 10 4 RLU, which was comparable to that of the background cell line that had not been virally infected (1 × 10 4 RLU; Figure 2). Therefore, the TZM-bl cell line was deemed an unsuitable host for our S-covered pseudo-typed lentivirus system.  We also examined LpVspike(+) infectivity in VeroE6/TMPRSS2 cells, which have previously been shown to be 10-fold more sensitive to infection than parental VeroE6 cells [16]. LpVspike(+) infected VeroE6/TMPRSS2 cells less efficiently compared to ACE2-CRFK and ACE2-HEK293T cells ( Figure 2). A reading of only 10 3 RLU was detected when VeroE6/TMPRSS2 cells were infected with 100 µL of viral solution. LpVspike(+) did not appear to infect ACE2-expressing Cos7 cells ( Figure 2).
The TZM-bl cell line (also called JC.53bl-13) is a HeLa cell derivative engineered to express CD4, CCR5 and CXCR4, as well as Tat-responsive firefly luciferase reporter genes. These features make TZM-bl cells highly susceptible to HIV-1 infection and enables quantification of viral infection and neutralizing antibody titers [17]. Because TZM-bl already has an integrated copy of the luciferase reporter gene, we constructed a second pseudo-typed lentivirus, named LpVspike2(+); this was essentially the same as LpVspike(+), except it did not contain the luciferase gene. ACE2-expressing TZM-bl cells infected with undiluted LpVspike2(+) produced a reading of 2 × 10 4 RLU, which was comparable to that of the background cell line that had not been virally infected (1 × 10 4 RLU; Figure 2). Therefore, the TZM-bl cell line was deemed an unsuitable host for our S-covered pseudo-typed lentivirus system.

The Comparison of LpVspike(+) Infectivity in HEK293T and CRFK Cell Lines with/without ACE2 Transfection
The S protein of SARS-CoV-2 uses ACE2 as its point of entry, and hence requires host ACE2 expression to initiate infection. We confirmed the specificity of this process by examining the infectivity of pseudo-typed lentiviruses containing the SARS-CoV-2 S protein (LpVspike(+)) and those that did not (LpVspike

The comparison of LpVspike(+) Infectivity in HEK293T and CRFK Cell Lines with/without ACE2 Transfection
The S protein of SARS-CoV-2 uses ACE2 as its point of entry, and hence requires host ACE2 expression to initiate infection. We confirmed the specificity of this process by examining the infectivity of pseudo-typed lentiviruses containing the SARS-CoV-2 S protein (LpVspike(+)) and those that did not (LpVspike
The addition of Deax increased the infection efficiency of both LpVspike(+) and LpVspike(−) 1000-fold in CRFK cells. While Deax increased the infection efficiency of LpVspike(+) 10-fold in ACE2-CRFK cells, it increased the infection efficiency of LpVspike(−) 1000-fold. These results suggest that Deax can cause non-specific infections in HEK293T, ACE2-HEK293T, CRFK, and ACE2-CRFK cell lines. Deax was therefore ruled out as a supplement for enhancing LpVspike infection efficiency for our system.

Effect of Polybrene on Infectivity
Hexadimethrine bromide, also known as polybrene, is a cationic polymer capable of enhancing in vitro HIV-1 infection by reducing electrostatic repulsion between virions and sialic acid on the cell surface [18]. Polybrene has been widely used in neutralization and infectivity assays examining HIV and pseudo-typed lentiviruses coated with the SARS-CoV-2 S protein [1,15,19,20]. Therefore, we examined the effect of polybrene in our pseudovirus system. Following LpVspike(+) infection, ACE2-CRFK cells produced a reading of approximately 10 4 RLU with 50 μL of viral solution both with and without polybrene supplementation, suggesting that polybrene did not provide any enhancement effect ( Figure 5b). For both CRFK and ACE2-CRFK cells, no LpVspike(−) infection was detected following polybrene addition (Figure 5a,b). Therefore, polybrene did not induce non-specific infections.

Effect of Polybrene on Infectivity
Hexadimethrine bromide, also known as polybrene, is a cationic polymer capable of enhancing in vitro HIV-1 infection by reducing electrostatic repulsion between virions and sialic acid on the cell surface [18]. Polybrene has been widely used in neutralization and infectivity assays examining HIV and pseudo-typed lentiviruses coated with the SARS-CoV-2 S protein [1,15,19,20]. Therefore, we examined the effect of polybrene in our pseudovirus system. Following LpVspike(+) infection, ACE2-CRFK cells produced a reading of approximately 10 4 RLU with 50 µL of viral solution both with and without polybrene supplementation, suggesting that polybrene did not provide any enhancement effect (Figure 5b). For both CRFK and ACE2-CRFK cells, no LpVspike(−) infection was detected following polybrene addition (Figure 5a,b). Therefore, polybrene did not induce non-specific infections.

Stability of ACE2 Expression in CRFK Cells and LpVspike(+)
We found that CRFK cells were the most sensitive of the cell lines tested, producing a 10-fold higher RLU reading compared to the HEK293T cells. In this experiment, ACE2 was expressed in CRFK cells using a transient system, as opposed to a stable genome-integrated system. We therefore set out to determine the stability of transient ACE2 expression in CRFK cells, with the goal of optimizing transfection conditions. Following seeding and harvesting, ACE2-transfected CRFK cells were incubated for three time periods: 1 h, 1 day, and 2 days. At each time point, the ACE2-transfected CRFK cells were exposed to LpVspike(+) in 96-well plates, then incubated for a further 48 h. The luciferase reporter measurements were similar for each time point, suggesting that the CRFK cells stably express the ACE2 receptor throughout the 48-h period following transfection with the ACE2 vector (Figure 6a).

Stability of ACE2 Expression in CRFK Cells and LpVspike(+)
We found that CRFK cells were the most sensitive of the cell lines tested, producing a 10-fold higher RLU reading compared to the HEK293T cells. In this experiment, ACE2 was expressed in CRFK cells using a transient system, as opposed to a stable genomeintegrated system. We therefore set out to determine the stability of transient ACE2 expression in CRFK cells, with the goal of optimizing transfection conditions. Following seeding and harvesting, ACE2-transfected CRFK cells were incubated for three time periods: 1 h, 1 day, and 2 days. At each time point, the ACE2-transfected CRFK cells were exposed to LpVspike(+) in 96-well plates, then incubated for a further 48 h. The luciferase reporter measurements were similar for each time point, suggesting that the CRFK cells stably express the ACE2 receptor throughout the 48-h period following transfection with the ACE2 vector (Figure 6a).
We then examined the stability of LpVspike(+) following incubation at 37 °C. After 1 h of incubation, luciferase expression had not changed, indicating that LpVspike(+) was stable (Figure 6b).

Inhibition of LpVspike(+) in ACE2-Expressing CRFK Cells with Plasma and Monoclonal Antibodies
We next assessed whether our LpVspike(+)/ACE2-CRFK cell system could be applied to assess antibody neutralization. We first calculated ID50 of two human convales- We then examined the stability of LpVspike(+) following incubation at 37 • C. After 1 h of incubation, luciferase expression had not changed, indicating that LpVspike(+) was stable (Figure 6b).

Discussion
In this study, we found that feline CRFK cells were 10-fold more sensitive to infection with S protein-coated HIV-1 pseudoviruses compared to human HEK293T cells. Given

Discussion
In this study, we found that feline CRFK cells were 10-fold more sensitive to infection with S protein-coated HIV-1 pseudoviruses compared to human HEK293T cells. Given that both pseudoviral S-protein expression and CRFK ACE2 expression were required for infection with LpVspike(+), we conclude that our system reflects the natural SARS-CoV-2 infection process. Moreover, this system was tested by measuring the neutralizing activities of convalescent sera and anti-spike mAbs, thus demonstrating its practicality.
In contrast to the high LpVspike(+) infectivity observed in CRFK cells, the infection rates of Cos7 and Vero cells, which are derived from macaque kidneys, were low. The LpVspike(+) pseudovirus was based on HIV-1, meaning that completion of the infection process following entry into the cytoplasm is dependent on whether the host cellular environment facilitates viral uncoating, reverse transcription, movement of the virus into the nucleus, and integration of the viral genome into chromosomes. Macaque cells possess restriction proteins such as APOBEC3 and TRIM5α/TRIM5CypA, which disrupt the HIV-1 genome and core [21], whereas cat cells express a truncated, non-functional version of Trim5α. CRFK cells may lack both functional APOBEC3 and Trim5α. Although human APOBC3 and Trim5α cannot prevent HIV-1 infection, human Trim5α exhibits weak restriction activity [22].
Our attempts to improve infection efficiency by adding Deax and polybrene were not successful. Deax caused non-specific infection in all combinations of cell types and pseudoviruses tested, regardless of the presence of the S protein or its ACE2 receptor. However, Deax specifically enhanced the infection of TZM-bl cells expressing human CD4/CCR5 with a HIV-1 envelope-coated pseudovirus. Although we do not have a definitive explanation for these results, we propose that Deax may disturb the membranes of both the cell and virus. Another unexpected result is that LpVspike(−) infected ACE2-HEK293T cells, but not ACE2-CRFK cells. ACE2 transfection might therefore perturb the membranes of HEK293T cells, causing non-specific infection.
The neutralization assay developed in this work using modified cat CRFK cells is an efficient and safe method that could be a useful research tool, enabling the examination of the binding mechanism between the SARS-CoV-2 S protein and the ACE2 receptor, as well as the development of antiviral drugs. As demonstrated above (Figure 8), this system could be employed to assess SARS-CoV-2 inhibitors and evaluate the neutralizing activity of convalescent antisera and anti-S protein mAbs; anti-S protein antibodies may be useful in treating critically ill COVID-19 patients. Although vaccine development is progressing at an unprecedented rate, this neutralization assay could still be a useful tool for determining the robustness of the B-cell response elicited by newly developed SARS-CoV-2 vaccines.

Neutralization Assays
ACE2-expressing CRFK cells were seeded into 100 µL of medium to a concentration of 2.5 × 10 5 cell mL −1 in 96-well plates 1 day before infection. Previously harvested samples containing pseudo-typed lentiviral particles coated with the SARS-CoV-2 S protein were thawed. To perform the virus titration, pseudo-typed virus samples were serially diluted two-fold a total of nine times and transferred to a 96-well plate. Control wells contained only supplemented medium and no virus sample. Each dilution was made in triplicate. For infection, 100 µL of each LpVspike(+) or LpVspike2(+) dilution was added to the cell-seeded 96-well plates, which were incubated at 37 • C under 5% CO 2 for 2 days.
Luciferase activity was measured in HEK293T, CRFK, Cos7, Vero, and TZM-bl cells. The details of the neutralization assays performed in this study have been described previously [26,27]. Briefly, to measure luciferase activity, 50 µL of cell lysate solution (Toyo B-Net, Tokyo, Japan) was added to each well, and the plate was agitated for 15 min. An aliquot of 30 µL of lysate was transferred to a Nunc F96 MicroWell white plate (Thermo Fisher Scientific, Waltham, MA, USA), and luminescent substrate (30 µL) was added to each well. Luciferase activity was measured using a TriStar LB 941 reader (Berthold Technologies, Bad Wildbad, Germany) and MikroWin software. ID50 values were calculated as previously described [28].
The neutralization assays were performed using adjusted viral doses that yielded equivalent infectivity levels for each cell line. The neutralization assays were performed in a 96-well format as described previously. ID50 was reported as either the concentration or sample dilution at which the RLU readout was reduced by 50% compared to the RLU readings measured in the virus control wells (cells plus virus without test sample) after subtracting background RLU values from cell control wells (cells only, no virus or test sample) [29].
To test the stability of ACE2-expression in CRFK cells, ACE2-expressing CRFK cells were seeded to a concentration of 2.5 × 10 5 mL −1 in 96-well plates and incubated for three different time periods: 1 h, 1 day and 2 days. Pseudo-typed viruses were then added to a final concentration of 6000 RLU mL −1 to the ACE2-expressing CRFK cells and incubated for 48 h at 37 • C under 5% CO 2 .
To evaluate the stability of LpVspike(+) when incubating at 37 • C, diluted pseudotyped viruses were incubated in triplicate for 10, 30 and 60 min. To test viral infectivity after incubation at 37 • C, virus samples were added to ACE2-expressing CRFK cells in 96-well plates and incubated for 48 h at 37 • C under 5% CO 2 .
SARS-CoV-2 plasma serum of COVID-19 patients with varying IgM and IgG antibody levels was purchased from RayBiotech (Peachtree Corners, GA, USA). The serum plasma was subjected to serial three-fold dilutions, from an initial dilution of 1:60 down to 1:1,180,980. We added 120 µL of supplemented medium to each well in a 96-well plate, with the exception of the outer edge wells, to which 250 µL of phosphate-buffered saline was added. Next, 180 µL of 1/20-or 1/40-diluted plasma was added to the first row of wells and subjected to three-fold serial dilutions. Control wells contained no plasma, only cells and virus. Then, 60 µL of pseudo-typed virus at a concentration of 6000 RLU mL −1 was added to wells containing the diluted plasma. The total volume of the diluted plasma and virus was approximately 180 µL. The plasma-virus mixtures were incubated at 37 • C under 5% CO 2 for 1 h. After incubation, 150 µL of the mixture was added to the seeded ACE2-expressing CRFK cells in 96-well plates. The plates were incubated at 37 • C under 5% CO 2 for 2 days, and then the luciferase activity of the samples was measured. ID50 was calculated as described previously [28].
Three anti-SARS-CoV-2 neutralizing mAbs were used, including human IgG (catalogue number E-AB-V1021), IgM (catalogue number E-AB-V1026), and IgA (catalogue number E-AB-V1027) isotypes with clone 8A5 Fab that was elicited using recombinant 2019-nCoV S-trimer Protein (His Tag) as the immunogen. The mAbs were diluted 10-fold three times from 1 to 0.001 µg mL −1 . The experimental method for testing neutralizing mAbs was the same as the method used for assessing the neutralizing ability of plasma. IC50 values were calculated as described previously [28]. N mAbs were purchased from Elabscience (Houston, TX, USA).

Conclusions
ACE2-expressing CRFK cells were 10 times more sensitive to viral infection compared to the typically used HEK293T cells. We tested our system with a neutralization assay and found that the pseudo-typed lentiviruses coated with SARS-CoV-2 S protein were neutralized by anti-SARS-CoV-2 IgG, IgM, and IgA mAbs and sera. We also evaluated the influence of Deax and polybrene on infectivity. Deax caused non-specific viral infection in the absence of ACE2 receptors and polybrene did not have a significant impact on infectivity or sensitivity. Our neutralization assay using pseudo-typed lentiviruses coated with SARS-CoV-2 S protein and ACE2-expressing CRFK cells is both easier and safer than the methods currently available and can aid researchers in understanding SARS-CoV-2 pathogenesis as well as discovering new drugs and vaccines.