Booster Gam-COVID-Vac vaccination elicits high titers of serum Ad26-neutralizing antibodies

Fifty-eight individuals were recruited to our study. All study participants were vaccinated with two injections of Gam-COVID-Vac in January–April, 2021. The first dose was rAd26-based, followed by the rAd5-based second dose 21 days later. After 9 months [range 249–300 days, Supplementary Table S1], all participants received their booster vaccination. In the core cohort (n = 48), rAd26-based first component of the Gam-COVID-Vac served as a booster, and in the comparison cohort (n = 10), the booster was Pfizer-BioNTech’s Comirnaty. Serum samples were collected at baseline (T0), one, three and six months later (T1, T3, and T6), and then immediately before, and one and four months after the booster immunization (T9, T10, and T14, respectively) (Fig. 1a).

Fig. 1: Booster vaccination with rAd26-vectored Gam-COVID-Vac elicits high titers of serum Ad26-specific NAbs.

a Timeline showing Gam-COVID-Vac vaccination and blood sampling timepoints, labeled according to the number of months elapsed after the first dose. b Outline of the rAd26 virus neutralization assay. In a subpanel, representative flow plots show Spike expression on the surface of A549 cells after rAd26-Spike infection in the absence of Ad26-neutralizing antibodies (left) and in the presence of a saturating amount of Ad26 Nab (right). Any anti-Spike activity in serum samples was blocked by adding the excess of recombinant Spike protein (not shown). c Representative rAd26 neutralization curves for serum samples before (T9) and one month after (T10) the Gam-COVID-Vac boost. Ad26 neutralizing activity was defined by the percentages of Spike+ A549 cells relative to no antibody controls. d Serum Ad26 virus-neutralizing antibody titers before and one month after booster vaccination in homologous (Gam-COVID-Vac/Gam-COVID-Vac) and heterologous (Gam-COVID-Vac/Pfizer-BioNTech’s Comirnaty) vaccination groups. Dotted lines indicate the pre-prime baseline. Medians are plotted and statistical significance was determined using Wilcoxon matched-pairs signed-rank test. Data are presented as median values and interquartile ranges (IQR). ****P < 0.0001.

The outline of the rAd26-neutralization assay used in our study is presented in Fig. 1b. A549 human lung carcinoma cells were used as the target for rAd26 infection. After 24 h of infection, approximately 50% of A549 cells expressed the full-length SARS-CoV-2 glycoprotein S encoded by the rAd26 vector, which was detected using anti-RBD monoclonal antibody XR15 (Fig. 1b, left panel). In the presence of saturating amount of anti-Ad26 NAbs, the infection was blocked, and no S expression was detectable (Fig. 1b, right panel). Sera obtained from the vaccinated individuals inhibited rAd26 vector entry in a concentration-dependent manner (Fig. 1c) and the level of anti-Ad26 NAbs was quantified as neutralization half-maximal inhibitory serum dilution (ID50). Usually, Ad NAb titers are assessed using rAd-luciferase or rAd-GFP reporter constructs21,22. However, we believe that the use of rAd directly from the Gam-COVID-Vac vaccine preparation provides more relevant information that is closer to the real-life situation.

First of all, we measured the titers of anti-vector antibodies before (T9) and after (T10) the booster Gam-COVID-Vac vaccination, in a core cohort with a homologous regimen of revaccination (i.e., in study participants who received Gam-COVID-Vac as both the prime and the boost). Sera from the vaccinated individuals sampled at a pre-boost timepoint T9 were found to noticeably inhibit infection of A549 cells with rAd26 vector (median = 124, Fig. 1d), with 79% (38/48) of the samples displaying anti-Ad26 NAb levels above the baseline (ID50 for healthy non-vaccinated donors, median = 50). These numbers are consistent with the fact that all study participants had their first dose of Gam-COVID-Vac, which is rAd26-based, 9 months earlier. Following revaccination, Ad26-NAb titers increased 4.2-fold (median ID50 = 523, P < 0.0001), compared to the pre-boost levels, and the anti-Ad26 seroprevalence reached 98%.

Individuals from the comparison arm received a heterologous boost (i.e., different platform/delivery vector/Spike). All participants who received an mRNA vaccine booster (n = 10), had unaltered Ad26 NAb levels (P = 0.3223). The levels of Ad5-specific NAbs remained the same in both homologous and heterologous vaccination groups (P = 0.0608; P = 0.1602, correspondingly) (Supplementary Fig. S1).

Thus, rAd26-based but not mRNA-based revaccination predictably led to the formation of anti-vector responses in all of the vaccinees. This is consistent with the idea that the increase in Ad26-neutralizing titers is due to the administration of rAd26-based vector, rather than the vaccination per se.

RBD-binding and SARS-CoV-2-neutralizing antibody titers are independent of the level of pre-boost anti-Ad26 antibodies

Of all Spike-specific antibodies, RBD-targeting antibodies are known to constitute the vast majority of SARS-CoV-2 NAbs2, hence in order to measure recall humoral responses after booster vaccination we first analyzed the levels of anti-RBD IgGs. Nine months after the prime and immediately before the booster shot (timepoint T9), study participants displayed pronounced RBD-specific IgG titers (median = 231 ng/mL). Seroprevalence of anti-RBD IgG was 60% (29/48). As expected, Gam-COVID-Vac revaccination resulted in a 2.5-fold increase in RBD-specific IgG levels (median = 582 ng/mL, P < 0.0001) and seroprevalence 94% (45/48) (Fig. 2a). An even more pronounced increase in RBD-specific IgGs was observed in the heterologous vaccination group (8.7-fold change, P = 0.0039) (Supplementary Fig. S2).

Fig. 2: Humoral anti-SARS-CoV-2 recall responses after booster Gam-COVID-Vac vaccination.

a Serum anti-RBD IgG levels before and one month after the booster injection measured by ELISA. b Spearman’s correlation between pre-boost (T9) Ad26-specific NAb values (x axis) and the serum levels of anti-RBD IgGs at T10 (y axis). Participant IDs are shown. c Schematic representation of the SARS-CoV-2 virus neutralization assay. d Neutralizing antibody titers (ID50) against SARS-CoV-2 WT, Alfa, Epsilon, Delta, and Omicron VOCs. Numbers indicate median values. e Spearman’s correlation between RBD-specific IgG and WT Spike SARS-CoV-2 NAbs at T10. f Lack of significant association between pre-boost (T9) Ad26-specific NAb levels with post-boost (T10) SARS-CoV-2 NAb titers against WT, Alfa, Epsilon, Delta, or Omicron VOCs. Numbers indicate Spearman’s rank correlation coefficients. g Comparison of RBD-binding IgG levels and SARS-CoV-2 Nab titers at T10 in subgroups with low (below the median) and high (above the median) Ad26 NAb ID50 at pre-boost timepoint T9. Medians are plotted and statistical significance was determined using Wilcoxon matched-pairs signed-rank test (a), Kruskal–Wallis test with Dunn’s post correction (d) or Mann–Whitney test (g).

To understand whether the level of pre-boost Ad26-specific immunity may affect the levels of post-boost Spike-specific antibodies, we tested for associations between the levels of RBD-specific IgGs at T10 vs. Ad26-specific NAbs at T9 (Fig. 2b). The Spearman’s correlation analysis did not reveal any association (Spearman r = −0.1855, P = 0.2068), indicating that pre-boost anti-Ad26 NAbs did not affect the production of RBD-binding IgGs.

While RBD-binding activity of antibodies is known to be important for protective immunity against SARS-CoV-2, their functional activity, i.e., the ability to neutralize the virus, may serve as a better predictor of protection. To quantify the SARS-CoV-2-neutralizing activity of sera from vaccinated individuals, virus neutralization assay was performed using lentiviral particles pseudotyped with Spike from SARS-CoV-2 WT (ancestral variant) or VOCs (Fig. 2c).

Before the booster at T9, serum samples demonstrated ID50 ~50 for WT, Alfa, and Epsilon VOCs of SARS-CoV-2. In addition, these sera also neutralized Delta and Omicron variant entry, albeit the latter was neutralized less effectively compared to the WT (P < 0.0001) (Fig. 2d).

Much as was observed for RBD-specific IgG responses, the WT SARS-CoV-2 neutralizing capacity of the sera significantly increased after the booster shot (T10) compared with the pre-booster (T9) levels (P = 0.0017). Furthermore, this was accompanied by higher neutralizing activities against Alfa, Epsilon, Delta, and Omicron VOCs (fold increase 4.0, 2.4, 2.3, and 3.9, respectively) (Supplementary Fig. S3). In the heterologous vaccination group receiving an mRNA-based boost, anti-SARS-CoV-2 NAb levels against WT, Alfa, Epsilon, Delta, and Omicron were even higher (20.9, 31.6, 19.1, 15.2, and 47.0-fold increase, respectively) (Supplementary Fig. S4).

We next sought for possible associations between the immunological parameters assayed. As expected, a good correlation between RBD-specific IgGs and WT Spike SARS-CoV-2 NAbs at T10 was observed (Spearman r = 0.8335, P < 0.0001) (Fig. 2e). At the same time, we failed to detect any association between the levels of Ad26-specific NAbs at T9 and SARS-CoV-2-neutralizing activities against WT or other tested VOCs (Alfa, Epsilon, Delta, and Omicron) (Spearman r = −0.1350, P = 0.3601; r = −0.1056, P = 0.4800; r = −0.0356, P = 0.8121; r = −0.0818, P = 0.5849; r = −0.1170, P = 0.4334, respectively, Fig. 2f, Supplementary Fig. S5). Next, the vaccinated individuals were stratified as having their Ad26 NAb ID50 values at the pre-boost timepoint T9 below or above the median. After booster vaccination (T10), the stratified subgroups did not differ in the levels of either RBD-binding IgG or SARS-CoV-2 NAb (P = 0.0948 and P = 0.1954, respectively, Fig. 2g). Taken together with the correlation analysis, these data suggest that both RBD-binding and SARS-CoV-2 NAb recall responses are independent of the pre-boost levels of Ad26-specific NAbs.

Dynamics of Ad26 virus-neutralizing, SARS-CoV-2 RBD-binding antibodies, and SARS-CoV-2 NAbs

Our analysis described above has established the increase in anti-Ad26 antibody titers following homologous revaccination. We asked when these antibodies appeared during the course of vaccinations and how long would these antibodies persist at detectable levels. In a subcohort of donors (n = 11), Ad26-specific NAb levels were monitored longitudinally from the first dose of the vaccine and up to 14 months at 7 timepoints. Figure 3a illustrates the dynamics of Ad26 neutralizing antibodies across six time intervals.

Fig. 3: Longitudinal analysis of humoral responses to Gam-COVID-Vac vaccination.

Ad26-specific NAb (a), Ad5-specific NAb (b), anti-RBD IgG (c), and WT SARS-CoV-2 NAb (d) responses before the first dose of Gam-COVID-Vac and up to 14 months later. LOWESS-smoothed lines are shown. Arrows indicate immunization timepoints. Asterisks indicate significant difference between groups determined using the Friedman test.

Pre-vaccine median level of Ad26 NAbs was on the baseline. It must be noted though that two individuals in our cohort had high pre-existing Ad26 NAb titers. One month after the prime vaccination, the levels of Ad26 NAbs demonstrated a marked increase, and slightly declined thereafter, yet persisted for the duration of 6 months of the follow-up. During the study, Ad26 NAb seroprevalence was invariably above 46%. Booster immunization restored Ad26 NAb levels back to the maximum levels, which was again followed by a minor decline. The magnitudes of Ad26 NAb peaks following prime and booster vaccination were generally comparable, and a 14.7- and 34.8-fold increase was observed relatively to the baseline (P = 0.0025, P < 0.0001, respectively). In contrast, the Ad5 response dynamics had a single peak soon after the prime Ad5-based vaccination (T1 vs. T0, 5.0-fold, P = 0.0007) and then gradually decreased to baseline at T14 (Fig. 3b).

The dynamics of RBD-specific IgG and WT SARS-CoV-2 NAb responses followed the pattern observed for Ad26 NAbs (Fig. 3c, d). Namely, after the second shot, RBD-binding IgG and WT SARS-CoV-2 NAb levels progressively declined for 6 months and reached a minimum before revaccination. Notably, RBD-binding antibody levels demonstrated the strongest decline, with nearly background levels found by the 9th month of the study. Nonetheless, WT SARS-CoV-2 NAb activity remained clearly detectable. Based on these numbers, setting a booster shot at 6–9 months after the prime appears well-substantiated, as this offers the advantage of declining Ad26 NAb titers with persisting SARS-CoV-2 humoral immunity.