Changes in COVID-19 antibody levels and types after multiple doses and different vaccine combinations

The study was conducted between January 2021 and January 2023 at the Institute of Medical Microbiology, Immunology and Parasitology (IMMIP), University Hospital Bonn. Participants were recruited in Bonn (Germany), Sremska Kamenica (Vojvodina, Serbia), and Banja Luka (Bosnia and Herzegovina), resulting in a total cohort of 331 individuals (99 men and 232 women). Samples from participants who received Gam-COVID-Vac or BBIBP-CorV vaccines were collected in Serbia as well as Bosnia and Herzegovina, since these vaccines have not been approved within the EU. All participants provided written informed consent before enrollment. Ethical approval was obtained from the University Hospital Bonn (approval code: 439/20) and the Faculty of Medicine, University of Novi Sad (approval code: FN.198/02). Epidemiological and clinical characteristics of the study population are summarized in Table 1. Venous blood samples were collected 2–4 weeks after the final vaccination dose. Individuals with immunodeficiencies or undergoing immunosuppressive therapy were excluded from the analyses. Infection-naive status was established using a combination of temporal, clinical, and serological criteria. First, baseline samples were collected early in the pandemic, prior to widespread community transmission. Participants underwent regular antigen self-testing and reported any symptoms potentially indicative of SARS-CoV-2 infection; only individuals with consistently negative tests and no history of symptoms were included. All participants were seronegative for Spike-specific IgA and IgG at baseline, indicating no prior exposure to SARS-CoV-2. Participants with previous infections were excluded from the analysis. All participants in the study received vaccines based exclusively on the original (wildtype) SARS-CoV-2 spike antigen. None of the individuals included in the analysis received variant-adapted vaccines. According to the vaccination records available for this cohort, all documented mRNA-1273 boosters were administered with the standard 50 µg dose recommended by the manufacturer. Reactogenicity was assessed using structured questionnaires, and reported side effects were classified according to duration. Severe reactions were defined as local or systemic symptoms—including injection-site pain, musculoskeletal complaints, fever, fatigue, or headache—lasting longer than three days, whereas mild reactions comprised the same symptoms resolving within three days. Samples from all 365 participants were collected after completion of the primary vaccination regimen (two doses). However, follow-up sampling at later time points, six months post–second dose and after booster vaccination, was not feasible for all individuals, resulting in varying sample sizes across the different time points.

The study evaluated six COVID-19 vaccines representing different technological platforms: two mRNA-based vaccines (mRNA-1273 and BNT162b2), two adenoviral vector–based vaccines (ChAdOx1-S, Gam-COVID-Vac), and one inactivated virus vaccine (BBIBP-CorV).All vaccines required two doses to complete the primary immunization. An overview of the vaccines assessed in this study, including their specific composition and underlying mode of action, is provided in Table 2.

SARS-CoV-2 spike-specific IgA and IgG antibodies were measured in plasma samples from European COVID-19 vaccine recipients using ELISA kits (Euroimmun, Lübeck, Germany; #EI2606-9601A and #EI2606-9601G). Samples were automatically diluted 1:101 in the provided buffer and incubated for 60 minutes at 37°C in pre-coated 96-well plates. The Euroimmun Analyzer I ELISA processor performed washing and incubation steps using the manufacturer's program. Optical density (OD) was determined at 450 nm, and results were classified as negative (<0.8), equivocal (0.8–1.1), or positive (>1.1).

SARS-CoV-2 neutralizing antibodies were quantified using the SARS-CoV-2 Variants Neutralizing Antibody 6-plex ProcartaPlex Panel (Thermo Fisher, Waltham, USA; #EPX060-16018-901) according to the manufacturer's instructions. This multiplex assay allows assessment of neutralizing capacity against six variants: Wildtype (WT), Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), Delta (B.1.617.2), and Omicron (B.1.1.529). The principle is based on magnetic beads conjugated with WT- or variant-specific proteins; neutralizing antibodies in the plasma compete with biotinylated ACE2 for binding, with streptavidin–phycoerythrin (PE) serving as the detection reagent. The signal is inversely proportional to the amount of neutralizing antibodies. In brief, 50 µL of magnetic beads were added to each well of a 96-well plate and washed with 150 µL of 1× wash solution. Prediluted plasma samples (25 µL, 1:100) and assay diluent (25 µL) were then added, alongside positive and negative controls. Plates were incubated for 2 h at room temperature (RT) with shaking (500 rpm). After two washes, 25 µL of 1× detection antibody was added and incubated for 30 min at RT with shaking. Following another two washes, 50 µL of streptavidin-PE solution was added and incubated for 30 min under the same conditions. Plates were washed twice, 120 µL of reading buffer was added, and samples were incubated for 5 min with shaking before measurement on a MagPix Luminex instrument. Neutralization was calculated as: Neutralization (%) = (1– MFI of sample/MFI of negative control) x 100. According to the manufacturer's guidelines, results ≥20% were considered positive for all variants except Omicron (B.1.1.529), for which the threshold was ≥25%.

High-binding 96-well plates (Greiner, Kremsmünster, Austria; #655061) were coated with SARS-CoV-2 Spike protein (Thermo Fisher, Waltham, USA; #RP-87700) diluted to 2 µg/ml in coating buffer (Thermo Fisher, #00-0044-59) at 50 µL per well and incubated overnight at 4°C. An arbitrary standard was generated from pooled plasma samples of three randomly selected COVID-19 vaccinees. Plates were washed three times with washing buffer and blocked with 200 µL/well ELISA buffer (Thermo Fisher, #00-4202-56) for 1 h at room temperature (RT). Following another washing step, diluted plasma samples (1:500 for IgG1; 1:200 for IgG3 and IgG4) were added (50 µL/well) and incubated overnight at 4°C. After washing, 50 µL/well of HRP-conjugated secondary antibodies were added – anti-human IgG1 (Thermo Fisher, #A-10648), anti-human IgG3 (Thermo Fisher, #05-3620), and anti-human IgG4 (Thermo Fisher, #MH1742) – and incubated for 2 h at RT. Plates were washed again, and 50 µL/well of TMB substrate (Thermo Fisher, #00-4201-56) was added and incubated for 15 min at RT. The reaction was terminated by adding 25 µL/well of stop solution (1 M H₂SO₄). Optical density (OD) was measured at 450 nm using a SpectraMax 190 reader (Molecular Devices, Munich, Germany). This approach enabled semiquantitative detection of SARS-CoV-2-specific IgG subclasses. As no cut-off values were defined, qualitative interpretation of antibody levels was not possible.

Data were analyzed using GraphPad Prism version 10.6.1. Participant characteristics were summarized as median ± IQR with minimum and maximum values for continuous variables, and as frequencies with percentages for categorical variables. Multiple group comparisons were performed using the nonparametric Kruskal–Wallis test followed by Dunn's post hoc test, while two-group comparisons were assessed with the Mann–Whitney U test. A p-value <no><0.05</no> was considered statistically significant.

Samples were collected at multiple time points: after the first dose, the second dose, six months following the second dose, and after the third (booster) dose. Participants were grouped based on their primary vaccination regimen (first and second doses) into the following categories: BNT162b2 (PZ), mRNA-1273 (MO), ChAdOx1-S (AZ), ChAdOx1-S/BNT162b2 (AZ/PZ), and ChAdOx1-S/mRNA-1273 (AZ/MO). Following abbreviations including: PZ, MO, AZ, GC, and SP were used throughout the manuscript. All individuals received an mRNA booster (either BNT162b2 or mRNA-1273) 6–8 months after their second dose. Following the first dose, significantly elevated levels of SARS-CoV-2-specific IgA and IgG antibodies were observed in the BioNTech, mRNA-1273, and ChAdOx1-S groups compared to the unvaccinated control group (Figure 1). Within these vaccine groups, the second dose resulted in a significant increase in IgG levels compared to the first dose, while IgA levels exhibited less pronounced changes. A similar pattern of significantly increased antibody levels after the second dose was also observed in the heterologous ChAdOx1-S/BNT162b2group (Figure 1). Six months after the second dose, a decline in both IgA and IgG antibody levels was observed across all vaccine groups, except the ChAdOx1-S vaccine. However, it should be noted that antibody levels in the ChAdOx1-S group did not exceed those observed in the mRNA-1273 and BNT162b2 groups at 6 months. Thus, despite a stronger decline in the latter, antibody responses remained at least comparable to those of ChAdOx1-S. Notably, antibody levels peaked after the third (booster) dose in all groups, regardless of the vaccine type received for the primary series. Each dot in Figure 1 represents the mean SARS-CoV-2-specific antibody level within each vaccine group. The fold-change and percentage increase or decrease in group mean antibody levels between the assessed time points (as depicted in Figure 1) are summarized in Table 3. These calculations include both IgA and IgG responses for all vaccine groups with available pre- and post-booster measurements, thereby providing a quantitative assessment of the magnitude and direction of booster-induced changes.

All participants received an mRNA booster (either BNT162b2or mRNA-1273) following a different primary two-dose vaccination regimen. SARS-CoV-2 spike-specific IgA and IgG antibody levels were compared after the booster vaccination, following different combinations of primary vaccinations. The booster dose led to a robust and statistically significant increase in both IgA and IgG antibody levels in all vaccinated individuals, compared to the unvaccinated control group—regardless of the combination of their initial two doses (Figures 2A, B). The median IgA levels were higher in the mRNA-1273 and BNT162b2 groups compared to those who received two homologous ChAdOx1-S doses or a mixed ChAdOx1-S/mRNA regimen. However, these differences were not statistically significant. Similarly, the third vaccine dose induced comparable IgG antibody levels across all groups, independent of their initial vaccination strategy (Figure 2B). The reactogenicity of the mRNA booster was assessed using a standardized questionnaire. Adverse reactions were categorized as asymptomatic, mild, or severe, based on the duration of the following symptoms: injection site pain, musculoskeletal symptoms, fever, fatigue, and headache. Mild symptoms were defined as the presence of at least one of these symptoms lasting less than three days, while severe symptoms were characterized by symptom duration exceeding three days after onset. A high proportion of individuals remained asymptomatic across the different vaccine groups. Nevertheless, most participants experienced mild side effects after receiving the booster (Figure 2C). No severe adverse reactions were reported in any group following the third dose, except for one case within the mRNA-1273 group (4.5%) (Figure 2C). Overall, reactogenicity varied slightly between the different COVID-19 vaccines. However, no vaccine-related serious adverse events requiring hospitalization were observed among study participants.

Following the observation of a strong and consistent antibody response after booster vaccination—regardless of the initial vaccine combination—the neutralizing capacity of SARS-CoV-2-specific antibodies against several variants of concern (VoCs) that emerged during the pandemic was further investigated. This was assessed using a Luminex™-based, cell- and virus-free ACE2 competition assay, which enables multiplex analysis in a single run. The tested VoCs included Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), Delta (B.1.617.2), and Omicron (B.1.1.529). High levels of neutralizing antibodies against all variants were detected in all groups after booster vaccination, independent of their primary (two-dose) vaccine regimen (Supplementary Figure 1). No statistically significant differences in neutralizing capacity were found between the groups. Additionally, the potential reduction in neutralization capacity against the VoCs compared to the original wildtype (WT) strain was examined across the different primary vaccination groups. Notably, comparable neutralizing antibody levels against all tested variants—including the WT—were observed within the BNT162b2, mRNA-1273, and ChAdOx1-S groups (Supplementary Figure 2). While only the ChAdOx1-S/mRNA-1273 (heterologous) group showed a statistically significant reduction in neutralization capacity against the Gamma (P.1) variant (Supplementary Figure 2), all vaccine groups exhibited a general trend toward decreased neutralization. However, this reduction did not reach statistical significance in the other groups.The neutralization assay provided both qualitative and quantitative insights, allowing evaluation of neutralizing antibody (NAb) seropositivity across different VoCs and at three distinct time points: after completion of the primary two-dose vaccination series (Figures 3A–D), six months post-second dose (Figures 3E–J), and following booster vaccination (Figures 3K–P). During the primary vaccination phase, various two-dose vaccine regimens were compared. Despite differences in the initial regimens, all participants later received an mRNA booster (BNT162b2 or mRNA-1273). The highest proportion of NAb-seronegative individuals after primary vaccination was observed among BBIBP-CorV recipients—both against the WT (Figure 3A) and especially against the VoCs (Figures 3B–D). Elevated rates of NAb seronegativity were also seen in the Gam-COVID-Vac and ChAdOx1-S groups, particularly against the Beta and Gamma variants (Figures 3C, D). In contrast, the lowest proportion of NAb seronegative individuals was observed in the mRNA (BNT162b2 or mRNA-1273) and heterologous (ChAdOx1-S/BNT162b2 or ChAdOx1-S/mRNA-1273) vaccine groups. At six months post-vaccination, a general decrease in NAb seropositivity was noted across all vaccine groups and variants, though to varying degrees (Figures 3E–J). The most significant loss of neutralizing activity was observed against the Omicron variant (Figure 3J). Interestingly, the neutralizing antibody seropositivity patterns at the second dose time points appear inconsistent. After the second dose (Figure 3A), the ChAdOx1-S/ChAdOx1-S group displayed lower NAb seropositivity rates compared with the heterologous ChAdOx1-S/BNT162b2 group, whereas at the six month follow-up (Figure 3E) this pattern was reversed, with a higher proportion of seropositive individuals in the ChAdOx1-S/ChAdOx1-S group. However, after administration of the booster dose, NAb seropositivity was restored in all vaccine groups and against all tested variants (Figures 3K, L).

Following the detection of SARS-CoV-2-specific IgG levels via ELISA and the assessment of neutralization potential using Luminex-based multiplex immunoassays, the distribution of IgG subclasses after multiple COVID-19 vaccinations was further examined. SARS-CoV-2-specific IgG1, IgG3, and IgG4 subclasses were quantified using ELISA. IgG subclass levels were measured at four key time points: after the first and second vaccine doses, six months after the second dose, and following the third (booster) dose. All individuals received an mRNA booster (BNT162b2 or mRNA-1273), after having completed either a homologous mRNA regimen (BNT162b2/BNT162b2 or mRNA-1273/mRNA-1273) or a heterologous combination of ChAdOx1-S followed by an mRNA vaccine. After booster vaccination, significantly elevated levels of SARS-CoV-2-specific IgG1 and IgG4 were observed across all groups, regardless of their initial vaccine combination (BNT162b2, mRNA-1273, or ChAdOx1-S/mRNA). IgG3 levels peaked after the second vaccine dose, particularly within the mRNA groups, with no significant change observed in the ChAdOx1-S/mRNA group across time points (Figure 4). Given the strong induction of IgG subclasses following the booster, further analysis of vaccine combinations revealed notable differences in the response to these subclasses. IgG1 and IgG3 levels were statistically similar across all groups—whether participants received two homologous mRNA doses (BNT162b2 or mRNA-1273) or a heterologous ChAdOx1-S/mRNA regimen (Supplementary Figures 3A, B). However, IgG4 levels showed a distinct pattern: participants who received two primary doses of mRNA-1273 exhibited significantly higher IgG4 responses compared to both the homologous BNT162b2 and heterologous ChAdOx1-S/mRNA groups (p = 0.0365 and p = 0.0014, respectively; Supplementary Figure 3C). Interestingly, while the composition of the primary vaccination series influenced IgG4 expression, the type of mRNA vaccine used for the booster (BNT162b2 vs. mRNA-1273) did not have a significant impact on subclass distribution (Supplementary Figures 3D–F). In addition to analyzing IgG subclass levels, we also examined the proportional distribution of of IgG1, IgG3, and IgG4 after vaccination with BNT162b2, mRNA-1273, and the heterologous ChAdOx1-S/mRNA regimen at four time points (Supplementary Figure 4). For both mRNA vaccines, IgG3 dominates after the first and second dose, while IgG1 and IgG4 represent smaller proportions. At six months, IgG3 remains the major subclass, but IgG1 and IgG4 increase slightly. After the third dose, the subclass pattern shifts toward higher proportions of IgG4 and IgG1 and a reduced contribution of IgG3. In the heterologous ChAdOx1-S/mRNA group, IgG3 is also predominant after the first two doses, with a rise in IgG1 over time. By the third dose, the distribution shifts more clearly toward IgG1, accompanied by decreased IgG3 and moderate IgG4 levels. However, it is important to note that IgG2 could not be detected and is therefore absent from the total IgG subclass distribution.