Measuring SARS-CoV-2 neutralizing antibody activity using pseudotyped and chimeric viruses

Abstract

The emergence of SARS-CoV-2 and the ensuing explosive epidemic of COVID-19 disease has generated a need for assays to rapidly and conveniently measure the antiviral activity of SARS-CoV-2-specific antibodies. Here, we describe a collection of approaches based on SARS-CoV-2 spike-pseudotyped, single-cycle, replication-defective human immunodeficiency virus type-1 (HIV-1), and vesicular stomatitis virus (VSV), as well as a replication-competent VSV/SARS-CoV-2 chimeric virus. While each surrogate virus exhibited subtle differences in the sensitivity with which neutralizing activity was detected, the neutralizing activity of both convalescent plasma and human monoclonal antibodies measured using each virus correlated quantitatively with neutralizing activity measured using an authentic SARS-CoV-2 neutralization assay. The assays described herein are adaptable to high throughput and are useful tools in the evaluation of serologic immunity conferred by vaccination or prior SARS-CoV-2 infection, as well as the potency of convalescent plasma or human monoclonal antibodies.

Graphical abstract

Figure 1.

Two-plasmid and three-plasmid HIV-1–based pseudotyped viruses. (A) Schematic representation of the modified HIV-1_NL ΔEnv-NanoLuc genome in which a deletion in env was introduced and Nef-coding sequences were replaced by those encoding a NanoLuc luciferase reporter. Infectious virus particles were generated by cotransfection of pHIV-1_NL4ΔEnv-NanoLuc and a plasmid encoding the SARS-CoV-2 S lacking the 19 amino acids at the C-terminus of the cytoplasmic tail (SΔ19). (B) Schematic representation of constructs used to generate SARS-CoV-2 S pseudotyped HIV-1–based particles in which HIV-1_NLGagPol, an HIV-1 reporter vector (pCCNanoLuc/GFP) encoding both NanoLuc luciferase and EGFP reporter, and the SARS-CoV-2 SΔ19 are each expressed on separate plasmids. RRE, HIV-1 Rev response element; WPRE, woodchuck hepatitis virus post-transcriptional regulatory element. (C) Infectivity measurements of HIV-1_NL ΔEnv-NanoLuc particles (generated using the plasmids depicted in A) on the indicated cell lines. Infectivity was quantified by measuring NanoLuc luciferase activity (RLUs) following infection of cells in 96-well plates with the indicated volumes of pseudotyped viruses. The mean and range of two technical replicates are shown. Target cells 293T/ACE2cl.22 and HT1080/ACE2cl.14 are single-cell clones engineered to express human ACE2 (see Fig. S1 A). Virus particles generated in the absence of viral envelope glycoproteins were used as background controls. (D) Same as C, but viruses were generated using the three plasmids depicted in B. (E) Infectivity measurements of CCNanoLuc/GFP containing SARS-CoV-2 pseudotyped particles generated using plasmids depicted in B on 293ACE2*(B) cells, quantified by measuring NanoLuc luciferase activity (RLU) or GFP levels (percentage of GFP-positive cells). Mean and range from two technical replicates are shown.

Figure S1.

Generation and HIV-1 pseudotype infection of ACE2-expressing cell lines. (A) 293T cells were stably transduced with a lentivirus vector CSIB expressing either wild-type ACE2 or catalytically inactive mutant ACE2*. Following selection, cells were used as uncloned bulk populations (B), or single-cell clones were isolated. Flow cytometry histograms show staining with an antibody against huACE2 (purple) or an isotype control (gray). (B) HT1080 cells were stably transduced as in A, and a single-cell clone used throughout this study is shown, stained as in A. (C) Infectivity of CCNanoLuc/GFP viruses, pseudotyped with either full-length or C-terminally truncated SARS-CoV and SARS-CoV-2 S proteins on 293T/ACE2*(B) cells. Virus particles generated in the absence of an S protein (No S) were used as background controls. Infectivity was quantified by measuring NanoLuc luciferase activity (RLU). Mean and range from two technical replicates is shown. (D) Infectivity of HIV-1_NLΔEnv-NanoLuc in the various cell lines. Virus generated in the absence of S is used as a background control and infectivity was quantified by measuring NanoLuc luciferase activity (RLU). Mean and range from two technical replicates are shown. (E) Same as D except that CCNanoLuc/GFP virus was used. (F) Effect of virus ultracentrifugation on the infectivity of HIV-1–based pseudotyped virus particles. 293T/ACE2*(B) cells were infected with equivalent doses of unconcentrated HIV-1_NLΔEnv-NanoLuc or the same virus that had be pelleted through 20% sucrose and then diluted to the original volume.

Figure S2.

Variables determining HIV-1 pseudotype infection signal. (A) Images (GFP) of confluent monolayers of the indicated cell lines after infection with equivalent amounts of CCNanoLuc/GFP pseudotyped with SARS-CoV-2 SΔ19 or no S protein, as indicated. Scale bar, 0.4 mm. (B) Relationship between NanoLuc luciferase activity (RLU) and ACE2 cell surface expression levels (quantified by flow cytometry; Fig. S1 A) following infection the cell lines depicted in Fig. S1 A with HIV-1_NLΔEnv-NanoLuc pseudotyped virus. Mean and range of two technical replicates are plotted. (C) Infectivity of CCNanoLuc/GFP pseudotyped virus generated by cotransfection with the indicated amounts of SARS-CoV-2 SΔ19 expression plasmid. Mean and range of two technical replicates are plotted. (D) Quantification of infectivity of HIV-1_NL4-3ΔEnv-NanoLuc pseudotype infection by immunostaining of 293T/ACE2(B) target cells with antibodies against the HIV-1 capsid (CA) protein. Nuclei were visualized by Hoechst staining. Scale bar, 70 µm.

Figure 2.

VSV-based SARS-CoV-2 pseudotyped viruses. (A) Schematic representation of the rVSVΔG/NG-NanoLuc genome in which G-coding sequences were replaced by an mNeonGreen-2A-NanoLuc luciferase reporter cassette. Infectious virus particles were generated by passaging G complemented rVSVΔG/NG-NanoLuc virus stocks through 293T cells transfected with a plasmid encoding SARS-CoV-2 SΔ19. (B) Infectivity of pseudotyped rVSVΔG/NG-NanoLuc particles on Huh7.5 cells was quantified by measuring luciferase activity (RLU) or the percentage of GFP-positive cells. Mean and range from two technical replicates are plotted. Virus particles generated by passage through cells that were not transfected with SARS-CoV-2 S were used as a control. (C) NanoLuc luciferase activity (RLU) in Huh7.5 cells measured at various times after infection with pseudotyped rVSVΔG/NG-NanoLuc particles. Mean and range from two technical replicates are plotted. (D) Infectivity of pseudotyped rVSVΔG/NG-NanoLuc particles on the indicated cell lines. Infectivity was quantified by measuring NanoLuc luciferase activity (RLU) following infection of cells in 96-well plates with the indicated volumes of pseudotyped viruses. Mean and range deviation from two technical replicates are shown.

Figure S3.

rVSVΔG/NG-NanoLuc pseudotyped virus infection. (A) Infection of Huh7.5 cells with the indicated volumes of rVSVΔG/NG-NanoLuc pseudotyped virus. Images of the entire well of 96-well plates are shown. (B) Infectivity of rVSVΔG/NG-NanoLuc pseudotyped with SARS-CoV-2 SΔ19 or no S (background control) on the indicated cell lines. Infectivity was quantified at the indicated times post-inoculation by measuring NanoLuc luciferase levels (RLU). Mean and range of two technical replicates are plotted. cl., clone. (C and D) HT1080-derived cell lines (C) or 293T-derived cell lines (D) were infected with varying amounts of rVSVΔG/NG-NanoLuc pseudotyped with SARS-CoV-2 SΔ19 or no S (background control), and NanoLuc luciferase levels were measured at 16 h after infection. Mean and range of two technical replicates are plotted. (E) Relationship between NanoLuc luciferase activity (RLU) and ACE2 cell surface expression levels (quantified by flow cytometry; Fig. S1 A) following infection the cell lines depicted in Fig. S1 A with rVSVΔG/NG-NanoLuc. Mean and range of two technical replicates is plotted. (F) Effect of virus concentration on the infectivity of rVSVΔG/NG-NanoLuc pseudotyped virus. Huh7.5 cells were infected with equivalent doses of either unmanipulated virus-containing supernatant or virions that had been pelleted by ultracentifugation or using Lenti-X and diluted to the original volume. Mean and range of two technical replicates are plotted.

Figure 3.

A replication-competent VSV/SARS-CoV-2 chimera. (A) Schematic representation of the rVSV/SARS-CoV-2/GFP genome in which G-encoding sequences were replaced by SARS-CoV-2 SΔ18 coding sequences. GFP-encoding sequences were introduced between the SARS-CoV-2 SΔ18 and L open reading frames. (B) Representative images of 293T/ACE2(B) cells infected with the indicated volumes of plaque-purified, adapted derivatives (2E1 and 1D7) of VSV/SARS-CoV-2/GFP following passage in the same cell line. Left and center images show contents of an entire well of a 96-well plate, and the right image shows an expanded view of the boxed areas containing individual plaques. (C) Infectivity measurements of rVSV/SARS-CoV-2/GFP virus stocks on 293T/ACE2(B) or control 293T cells, quantified by measuring the percentage of GFP-positive cells at 16 h after infection. Mean and range from two technical replicates are shown. (D) Schematic representation of the adaptive changes acquired in rVSV/SARS-CoV-2/GFP during passage. Changes in 1D7 and 2E1 are shown in blue and red, respectively. PRRAR, amino acid sequence at the furin cleavage site; TM, transmembrane domain.

Figure S4.

Examples of neutralization of HIV-1 and VSV pseudotyped virus particles by mAbs targeting SARS-CoV-2 S. (A) Images of Huh7.5 cells following infection with rVSVΔG/NG-NanoLuc pseudotyped virus (∼10³ IU/well) in the presence of the indicated concentrations of a human mAb (C144) targeting SARS-CoV-2 S RBD. Images of the entire well of a 96-well plate are shown. (B) Quantification of rVSVΔG/NG-NanoLuc pseudotyped virus infection (measured by flow cytometry (percentage of mNeonGreen-positive cells, green) or by NanoLuc luciferase activity (RLU, blue) in the presence of the indicated concentrations of a human mAb (C102) targeting SARS-CoV-2 S RBD or a control mAb against the Zika virus envelope glycoprotein. Mean and range of two technical replicates are plotted. (C) Quantification of HIV-1_NLΔEnv-NanoLuc or CCNanoLuc/GFP pseudotyped virus infection on the indicated cell lines in the presence of the indicated concentrations of a human mAb (C121) targeting SARS-CoV-2 S RBD infectivity was quantified by measuring NanoLuc luciferase levels (RLU). Mean and range of two technical replicates are plotted.

Figure 4.

Measurement of neutralization activity in COVID-19 convalescent donor plasma. (A) Plasma neutralization of SARS-CoV-2. Serial fivefold dilutions of plasma samples from convalescent donors were incubated with SARS-CoV-2 (n = 3 replicates) and residual infectivity determined using VeroE6 target cells, expressed as percentage of infected cells by immunostaining. (B) Plasma neutralization of HIV-1_NLΔEnv-NanoLuc pseudotyped virus using 293T/ACE2*(B) target cells, rVSVΔG/NG-NanoLuc pseudotyped virus using Huh7.5 target cells, or replication-competent rVSV/SARS-CoV-2/GFP using 293T/ACE2(B) target cells. Residual infectivity was quantified by measuring either NanoLuc luciferase (RLU) or the percentage of GFP-positive cells, as indicated. (C) Correlation between NT₅₀ values for each of the 20 plasmas for each of the surrogate viruses (x axis) and NT₅₀ values for the same plasmas for SARS-CoV-2 (y axis).

Figure 5.

Measurement of neutralization potency of human mAbs. (A) Neutralization of SARS-CoV-2. The indicated concentrations of mAbs were incubated with SARS-CoV-2 (n = 3 replicates) and residual infectivity determined using Vero E6 target cells (expressed as the percentage of infected cells) by immunostaining. (B) mAb neutralization of HIV-1_NLΔEnv-NanoLuc pseudotyped virus using 293T/ACE2*(B) target cells, rVSVΔG/NG-NanoLuc pseudotyped virus using Huh7.5 target cells, or replication-competent rVSV/SARS-CoV-2/GFP using 293T/ACE2(B) target cells. Residual infectivity was quantified by measuring either NanoLuc luciferase (RLU) or the percentage of GFP-positive cells, as indicated. (C) Correlation between IC₅₀ values for each of the 15 mAbs for each of the surrogate viruses (x axis) and IC₅₀ values for the same antibodies for SARS-CoV-2 (y axis).