Comparative analysis of the neutralizing activity against SARS-CoV-2 Wuhan-Hu-1 strain and variants of concern: Performance evaluation of a pseudovirus-based neutralization assay

Abstract

Objectives: Emergence of new variants of SARS-CoV-2 might affect vaccine efficacy. Therefore, assessing the capacity of sera to neutralize variants of concern (VOCs) in BSL-2 conditions will help evaluating the immune status of population following vaccination or infection.

Methods: Pseudotyped viruses bearing SARS-CoV-2 spike protein from Wuhan-Hu-1/D614G strains (wild type, WT), B.1.617.2 (Delta), or B.1.1.529 (Omicron) VOCs were generated to assess the neutralizing antibodies (nAbs) activity by a pseudovirus-based neutralization assay (PVNA). PVNA performance was assessed in comparison to the micro-neutralization test (MNT) based on live viruses. Sera collected from COVID-19 convalescents and vaccinees receiving mRNA (BNT16b2 or mRNA-1273) or viral vector (AZD1222 or Ad26.COV2.S) vaccines were used to measure nAbs elicited by two-dose BNT16b2, mRNA-1273, AZD1222 or one-dose Ad26.CO2.S, at different times from completed vaccination, ~ 1.5 month and ~ 4-6 months. Sera from pre-pandemic and unvaccinated individuals were analyzed as controls. Neutralizing activity following booster vaccinations against VOCs was also determined.

Results: PVNA titers correlated with the gold standard MNT assay, validating the reliability of PVNA. Sera analyzed late from the second dose showed a reduced neutralization activity compared to sera collected earlier. Ad26.CO2.S vaccination led to very low or absent nAbs. Neutralization of Delta and Omicron BA.1 VOCs showed significant reduction of nAbs respect to WT strain. Importantly, booster doses enhanced Omicron BA.1 nAbs, with persistent levels at 3 months from boosting.

Conclusions: PVNA is a reliable tool for assessing anti-SARS-CoV-2 nAbs helping the establishment of a correlate of protection and the management of vaccination strategies.

Keywords: COVID-19; SARS-CoV-2; immunogenicity; neutralization assay; neutralizing antibodies; pseudotyped virus; vaccines; variants of concern.

Figure 1

Serum collection schedule and nAb antibody detection. (A) Schematic representation of time points (blue: short-time; orange: long-time) of serum collection for each cohort of donors. The median time from the 2nd dose of BNT162b2, mRNA-1273, and AZD1222 was of 56 days (range 20–68 days), 44.5 days (range 22–67), and 55.5 days (range 20–72), respectively. The median time from the 2nd dose of BNT162b2, mRNA-1273, AZD1222, and from the 1st dose of Ad26.CO2.S was of 138 days (range 101–191 days), 128 days (range 110–172), 132 days (range 110–192), and 136 days (range 120–162), respectively. (B) 90% live SARS-CoV-2 wild type (WT) neutralization titers (MNT₉₀, left y-axis, hatched bars) and 90% pseudovirus neutralization titers (pVNT₉₀, right y-axis, open bars) were used as measure of nAb levels in sera from SARS-CoV-2 infected recovered donors or sera from vaccinated donors collected at different time points. Geometric means ± 95% confidence intervals were reported. Neutralization titers < 10 were considered negative and given an arbitrary value of 5 (dotted line). Kruskal–Wallis test with Dunn’s multiple comparison post-test was used to compare groups. ** p ≤ 0.01; **** p < 0.0001. (C) Correlation between pVNT₉₀ and MNT₉₀ titers against SARS-CoV-2 WT strain. The non-parametric Spearman’s correlation coefficients (r_s), statistically significant p values and ratio between pVNT₉₀ mean titers and MNT₉₀ mean titers were provided. Perfect-fit correlation line was included on the plot. GMT: geometric mean titers; n = number of samples.

Figure 2

Live virus MNT and SARS-CoV-2 PVNA on sera collected from vaccinated donors produce comparable results. (A, B) 90% live SARS-CoV-2 wild type (WT) neutralization titers (MNT₉₀, left y-axis, hatched bars) and 90% pseudovirus neutralization titers using pseudotyped lentiviruses carrying the WT SΔ19 protein (pVNT₉₀, right y-axis, open bars) in sera drawn from (A) unvaccinated and recipients of two-dose BNT162b2, mRNA-1273, and AZD1222 collected at a median time of ~ 1.5 mo post-second vaccine dose; (B) unvaccinated and recipients of two-dose BNT162b2, mRNA-1273, AZD1222, or one dose Ad26.CO2.S collected at a median time of ~ 4–6 mo post-vaccination. Geometric means ± 95% confidence intervals were reported. Neutralization titers < 10 were considered negative and given an arbitrary value of 5 (dotted line). Kruskal–Wallis test with Dunn’s multiple comparison post-test was used to compare groups. * p < 0.05; ** p ≤ 0.01; *** p ≤ 0.001; **** p < 0.0001. (C) Correlation between pVNT₉₀ and MNT₉₀ titers shown in (A); (D) correlation between pVNT₉₀ and MNT₉₀ titers shown in (B). The non-parametric Spearman’s correlation coefficients (r_s), statistically significant p values and ratio between pVNT₉₀ mean titers and MNT₉₀ mean titers were provided. Perfect-fit correlation line was included on the plots. GMT: geometric mean titers; n = number of samples.

Figure 3

Neutralizing antibodies against SARS-CoV-2 Delta variant are detected either by MNT or PVNA. (A, B) 90% live SARS-CoV-2 Delta neutralization titers (MNT₉₀, left y-axis, hatched bars) and 90% pseudovirus neutralization titers using pseudotyped lentiviruses carrying the Delta SΔ19 protein (pVNT₉₀, right y-axis, open bars) in sera drawn from (A) unvaccinated and recipients of two-dose BNT162b2, mRNA-1273, and AZD1222 collected at a median time of ~ 1.5 mo post-second vaccine dose; (B) unvaccinated and recipients of two-dose BNT162b2, mRNA-1273, AZD1222, or one dose Ad26.CO2.S collected at a median time of ~ 4–6 mo post-vaccination. Geometric means ± 95% confidence intervals were reported. Neutralization titers < 10 were considered negative and given an arbitrary value of 5 (dotted line). Kruskal–Wallis test with Dunn’s multiple comparison post-test was used to compare groups. * p < 0.05; ** p ≤ 0.01; *** p ≤ 0.001. (C) Correlation between pVNT₉₀ and MNT₉₀ titers shown in (A); (D) correlation between pVNT₉₀ and MNT₉₀ titers shown in (B). The non-parametric Spearman’s correlation coefficients (r_s), statistically significant p values and ratio between pVNT₉₀ mean titers and MNT₉₀ mean titers were provided. Perfect-fit correlation line was included on the plots. GMT: geometric mean titers; n = number of samples.

Figure 4

Neutralizing antibodies against SARS-CoV-2 Omicron variant are detected either by MNT or PVNA. (A, B) 90% live SARS-CoV-2 Omicron neutralization titers (MNT₉₀, left y-axis, hatched bars) and 90% pseudovirus neutralization titers using pseudotyped lentiviruses carrying the Omicron SΔ19 protein (pVNT₉₀, right y-axis, open bars) in sera drawn from (A) unvaccinated and recipients of two-dose BNT162b2, mRNA-1273, and AZD1222 collected at a median time of ~ 1.5 mo post-second vaccine dose; (B) unvaccinated and recipients of two-dose BNT162b2, mRNA-1273, AZD1222, or one dose Ad26.CO2.S collected at a median time of ~ 4–6 mo post-vaccination. Geometric means ± 95% confidence intervals were reported. Neutralization titers < 10 were considered negative and given an arbitrary value of 5 (dotted line). Kruskal–Wallis test with Dunn’s multiple comparison post-test was used to compare vaccine groups. * p < 0.05. (C) Correlation between pVNT₉₀ and MNT₉₀ titers shown in (A); (D) correlation between pVNT₉₀ and MNT₉₀ titers shown in (B). The non-parametric Spearman’s correlation coefficients (r_s), statistically significant p values and ratio between pVNT₉₀ mean titers and MNT₉₀ mean titers were provided. Perfect-fit correlation line was included on the plots. GMT: geometric mean titers; n = number of samples.

Figure 5

Vaccine booster shots strengthen nAb response against Omicron VOC. (A) Schematic representation of serum collection time points. (B, C) Violin plots representing Log of pVNT₉₀ observed in sera (n = 15) collected at ~ 1 mo or ~ 3 mo after vaccine boost and challenged with (B) SARS-CoV-2 WT or (C) Omicron pseudotyped lentiviruses. A two-tailed non-parametric Mann–Whitney test for unpaired observations was performed. * p < 0.05; ** p ≤ 0.01. (D) pVNT₉₀ titers against Omicron VOC in sera collected at ~ 1 mo and ~ 3 mo after vaccine boost report immune response waning after three months. ** p ≤ 0.01. (E) pVNT90 titers against Omicron VOC observed in sera from three donors infected with SARS-CoV-2 and excluded from the previous analyses. A two-tailed non-parametric Wilcoxon signed-rank test was performed for paired observations (D, E). Geometric mean titers are reported. n = number of samples.