Study design and population
In this prospective longitudinal study, PLWH receiving either an original–BA.4/5 bivalent booster (Comirnaty® Original/Omicron BA.4/5, Pfizer-BioNTech) or an original monovalent booster (Comirnaty®, Pfizer-BioNTech) as a fourth vaccine dose were consecutively enrolled at the Clinic of Infectious Diseases and Tropical Medicine, San Paolo Hospital, ASST Santi Paolo e Carlo, Department of Health Sciences, University of Milan (Milan, Italy), between May 2022 and February 2023, and followed-up until a period between November 2022 and November 2023.
HIV-negative HCWs receiving the Comirnaty® Original/Omicron BA.4/5 booster were also enrolled in the same period at the Occupational Medicine Unit, Verona Hospital, Department of Diagnostics and Public Health, University of Verona, Verona, Italy.
Peripheral blood samples were collected in EDTA tubes from all participants at three different time-points: the day of the booster administration (T0), 1 month after the booster (T1), and 4–9 months after the booster (T2; only for PLWH and HCWs who received the original–BA.4/5 bivalent booster) (Fig. 1a).
Schematic representation of the study design showing the time-points at which study participants were sampled (a), as well as the number of participants who were enrolled and the immune parameters that were measured at each specified time-point to explore: b peak and durability of immune responses to the original–BA.4/5 bivalent booster in PLWH either with or without previous SARS-CoV-2 infection; c the possible role of the ancestral spike in the bivalent vaccine in exacerbating immunologic imprinting thus affecting the bivalent booster immunogenicity in PLWH; d possible differences in bivalent booster immunogenicity and durability between PLWH and HIV-negative HCWs. Created in BioRender [Augello, M. (2025), https://BioRender.com/t94l162].
Demographic, clinical, and HIV-related characteristics of the study participants were collected at enrollment.
SARS-CoV-2 infection was determined by anamnesis and/or anti-nucleocapsid serology at baseline and during the follow-up period.
To determine peak and durability of immune responses to the original–BA.4/5 bivalent booster in PLWH, SARS-CoV-2–specific T and B cells as well as humoral immunity (RBD-binding and RBD-blocking antibodies) were measured at all time-points and then analyzed both overall and according to previous SARS-CoV-2 infection (Fig. 1b). The role of the HIV-related dysfunction in shaping immune responses to the booster was explored by computing the association between baseline (T0) T cell phenotypes and peak response (fold-increase, T1/T0 ratio) or waning (fold-decrease, T1/T2 ratio) of vaccine-induced cellular/humoral responses adjusting for potential confounders.
To explore the possible role of the ancestral spike in the bivalent vaccine in exacerbating immunologic imprinting thus affecting the bivalent booster immunogenicity, fold-increase of vaccine-elicited humoral immunity was compared between SARS-CoV-2–naïve PLWH vaccinated with the original–BA.4/5 booster and PLWH (without previous SARS-CoV-2 infection) who received the original monovalent booster (Fig. 1c).
Lastly, to assess whether the bivalent booster immunogenicity and durability may be lower in PLWH compared to the general population, vaccine-induced antibodies fold-increase and fold-decrease were determined in HIV-negative HCWs and compared to those of PLWH adjusting for potential confounders (Fig. 1d).
The study was approved by the local Ethics Committee (Comitato Etico Milano Area 1, protocol number 0002574, 20/01/2022) and informed consent was obtained from each participant. All research was performed in accordance with the Declaration of Helsinki.
Plasma and peripheral blood mononuclear cells isolation
Peripheral blood samples from each participants were used to isolate plasma and peripheral blood mononuclear cells (PBMCs). Plasma was separated by centrifugation and stored at –80 °C until use. PBMCs were obtained by Ficoll density gradient centrifugation (Biowest Lymphosep, Lymphocyte Separation Media, Cat# L0560) and stored at –80 °C and then in liquid nitrogen until use.
Markers of HIV-related T cell dysfunction
T cell phenotypes were quantified at baseline (T0) by flow cytometry. Briefly, flow cytometry samples were prepared with the automated BD FACSDuet™ Sample Preparation System by mixing 50 μL of fresh whole blood and 40 μL of fluorescently-labeled antibodies cocktail (CD45–V500, CD4–APC-Cy7, CD8–PerCP-Cy5.5, CD127–PE, CD38–APC, CD45RA–PE-Cy7) in BD Trucount™ Tubes. After a 20-minute incubation at room temperature, red blood cells were lysed with BD FACS™ Lysing Solution. Samples were then acquired on a BD FACSLyric™ Flow Cytometry System, and fcs files analyzed with BD FACSuite™ Clinical Application software. The gating strategy for a representative sample (generated with FlowJo v10) is shown in Supplementary Fig. 1. The following phenotypes were determined within the lymphocyte gate and quantified as cells/μL: CD4 T cells (CD4+), CD8 T cells (CD8+), functionally-competent CD4 T cells (CD4+CD127+), activated CD8 T cells (CD8+CD38+), naïve CD4 T cells (CD4+CD45RA+), and naïve CD8 T cells (CD8+CD45RA+).
Anti-nucleocapsid serology
Anti-nucleocapsid (N) IgG were measured by means of a commercial kit [EUROIMMUN Anti-SARS-CoV-2 NCP ELISA IgG (Cat# EI 2606-9601-2G)] according to manufacturer’s instructions. Briefly, EDTA plasma samples were diluted 1:101 in dilution buffer. 100 μL of diluted samples, calibrator, positive control, and negative control were loaded in duplicate into nucleocapsid-coated wells and incubated for 1 hour at +37 °C. Plates were washed ×3, and then incubated with 100 μL/well of peroxidase-conjugated anti-human IgG antibodies for 30 minutes at room temperature (RT). After washing ×3, plates were incubated with 100 μL/well of chromogen substrate [tetramethylbenzidine (TMB)/H2O2] for 15 minutes at RT in the dark, and then the reaction was quenched with 100 μL/well of stop solution (H2SO4 0.5 M). The optical density (OD) was measured immediately by using PerkinElmer EnSight™ Multimode Plate Reader at 450 nm and 650 nm. Samples were considered positive for anti-N IgG antibodies by means of a semi-quantitative method, according to the following formula: [(sample OD450 – sample OD650) / (calibrator OD450 – calibrator OD650)] ≥1.1.
SARS-CoV-2–specific T cells
SARS-CoV-2–specific T cells reactive against WT and BA.4/5 virus were assessed in cryopreserved PBMCs by means of intracellular cytokine staining (ICS) and degranulation assay in flow cytometry. Briefly, 1.5 ×106 thawed PBMCs were plated in RPMI 1640 (EuroClone RPMI, Cat# ECB2000) supplemented with 1% L-glutamine (Gibco, Cat# 25030-149), 1% penicillin-streptomycin (Sigma, Cat# P4333-100ML), and 10% human serum, and rested for 3 hours at +37 °C in CO2 incubator. PBMCs were then stimulated for 18 hours with overlapping peptide pools derived from the WT and BA.4/5 spike protein (Miltenyi Biotec PepTivator® SARS-CoV-2 Prot_S B.1.1.529/BA.5 WT Reference Pool – research grade, Cat# 130-132-050, and Miltenyi Biotec PepTivator® SARS-CoV-2 Prot_S B.1.1.529/BA.5 Mutation Pool – research grade, Cat# 130-132-051) at a concentration of 1 μg/mL. Staphylococcus aureus enterotoxin B (SEB) (Sigma-Aldrich, Cat# S4881-1MG) at a concentration of 1 μg/mL was used as positive control, while negative controls were left untreated. CD107a–PE antibody (BD Pharmingen™, Cat# 555801) was also added to each well for the degranulation assay. After 1 hour of stimulation brefeldin A [BD GolgiPlug™ Protein Transport Inhibitor (Containing Brefeldin A), Cat# 555029] and monensin [BD GolgiStop™ Protein Transport Inhibitor (Containing Monensin), Cat# 554724] were added to wells. After stimulation, cells were harvested, washed, and stained with viability dye [Viobility Fixable Dye 405/520 (Miltenyi Biotec, Cat# 120-028-574)] for 15 minutes at room temperature (RT) in the dark. After a washing step, cells were stained for surface markers for 30 minutes at +4 °C in the dark with the following fluorescently-labeled antibodies: CD3–FITC (BD Pharmingen™, Cat# 555339), CD4–APC-Vio®770 (Miltenyi Biotec, Cat# 130-113-223), CD8–PerCP-Vio®700 (Miltenyi Biotec, Cat# 130-110-682). After washing, cells were fixed and permeabilized with fixation/permeabilization solution (BD Cytofix/Cytoperm™ Fixation/Permeabilization Kit, Cat# 554714) for 20 minutes at +4 °C, then further washed and stained for intracellular cytokines for 30 minutes at +4 °C with the following fluorescently-labeled antibodies: IFN-γ–PE-Cy™7 (BD Pharmingen™, Cat# 557643), TNF-α–V450 (BD Horizon™, Cat# 561311), and IL-2–APC (Miltenyi Biotec, Cat# 130-111-304). Cells were then washed and resuspended in PBS. Samples were acquired using FACSVerse™ cytometer (BD Biosciences) and FCS files were analyzed using FlowJo 10.9 (BD Biosciences).
SARS-CoV-2–specific T cells were determined subtracting unspecific background cytokines (IFN-γ, TNF-α, IL-2) production or degranulation (CD107a) in paired unstimulated samples (negative control) from stimulated samples; negative values were set to zero. SARS-CoV-2–specific Th1 and Tc1 cells were determined using the FlowJo Boolean OR Gate to identify CD4 and CD8 T cells expressing at least one Th1-like cytokine (IFN-γ, TNF-α, or IL-2). T cell polyfunctionality was assessed by using the FlowJo Boolean Combination Gates and SPICE 6.0 to identify single-, dual-, triple- cytokine-producing SARS-CoV-2–specific Th1/Tc1 cells. Representative plots of the gating strategy are shown in Supplementary Fig. 2.
SARS-CoV-2–specific B cells
SARS-CoV-2–specific B cells reactive against WT and BA.4/5 virus were measured in cryopreserved PBMCs by means of a commercial kit (Miltenyi Biotec SARS-CoV-2 RBD B Cell Analysis Kit, anti-human, Cat# 130-128-032), according to manufacturer’s instruction with few modifications. This assay is based on a double discrimination method to identify and phenotype RBD-specific B cells in flow cytometry after staining with RBD fluorescently-labeled tetramers. Firstly, peripheral blood mononuclear cells (PBMCs) were thawed and rested in RPMI 1640 (EuroClone RPMI, Cat# ECB2000) supplemented with 1% L-glutamine (Gibco, Cat# 25030-149), 1% penicillin-streptomycin (Sigma, Cat# P4333-100ML), and 10% human serum for 2 hours at +37 °C in CO2 incubator. In the meanwhile, RBD-tetramer solutions were prepared by mixing 15 μg/mL biotinylated WT-RBD [Recombinant SARS-CoV-2 RBD (HEK)-Biotin (provided with the kit)] or biotinylated BA.4/5-RBD [ACROBiosystems Biotinylated SARS-CoV-2 Spike RBD, His,Avitag™ (BA.4&BA.5/Omicron), Cat# SPD-C82Ew] with streptavidin–PE (for RBD-tetramer-PE) or streptavidin–PE-Vio770 (for RBD-tetramer-PE-Vio770) in PEB buffer [PBS + 0.5% BSA (Sigma-Aldrich, Cat# A7030-100G) + EDTA 2 mM (EMD Millipore, Cat# 324504-500 ML)] for 15 minutes at room temperature (RT). 5–10 ×106 thawed PBMCs were then incubated with the antibody staining cocktail for 30 minutes at +4 °C: WT-RBD-tetramers or BA.4/5-RBD-tetramers, 7-AAD Staining Solution, CD19–APC-Vio770, CD27–VioBright-FITC, IgM–APC, IgG–VioBlue®, IgA–VioGreen™. After washing, cell pellet was resuspended in PEB buffer. Samples were acquired using FACSVerse™ cytometer (BD Biosciences) and FCS files were analyzed using FlowJo 10.9 (BD Biosciences). Memory B cells were defined as CD19+CD27+ live cells. SARS-CoV-2–specific memory B cells were identified by the double discrimination method (RBD-tetramer-PE+RBD-tetramer-PE-Vio770+) within memory B cells. IgM+, IgG+, and IgA+ cells where then gated within SARS-CoV-2–specific memory B cells. Representative plots of the gating strategy are shown in Supplementary Fig. 3.
RBD-binding antibodies
Total RBD-binding antibodies reactive against WT and BA.4/5 virus were measured in EDTA plasma samples by a homemade ELISA16,20,32. Briefly, high-binding 96-well plates (Greiner Bio-One, Cat# 655061) were coated with 3 µg/mL of recombinant wild-type SARS-CoV-2 receptor binding domain (RBD) [ACROBiosystems SARS-CoV-2 (COVID-19) S protein RBD, His Tag (MALS verified), Cat# SPD-C52H3] or recombinant Omicron BA.4/5 SARS-CoV-2 RBD [ACROBiosystems SARS-CoV-2 Spike RBD Protein, His Tag (BA.4&BA.5/Omicron) (MALS verified), Cat# SPD-C522r] diluted in 50 mM of carbonate-bicarbonate buffer pH 9.6 (Sigma-Aldrich, Cat# C3041-50cap) and incubated overnight at +4 °C. Plates were then washed with washing buffer [PBS + 0.05% TWEEN®20 (Sigma, Cat# P9416-100ML)] and blocked for 1.5 hours with blocking buffer [PBS + 2% bovine serum albumin (BSA) (Sigma-Aldrich, Cat# A7030-100G)] at +37 °C. Plasma samples were serially diluted in assay buffer (PBS + 1% BSA) (1:200, 1:1200, 1:7200), loaded (50 μL/well) into plate wells in duplicates, and incubated for 2 hours at +37 °C. After washing ×3, a mix of biotinylated goat anti-human k and λ light chain antibodies (Bethyl Laboratories, Cat# A80-115B and A80-116B) were used at 1:2500 dilution for detection, followed by avidin-HRP (ThermoFisher, Cat# 18-4100-51) diluted at 1:2000, for 30 minutes at RT in the dark and mild agitation. After washing ×3, the detection was carried out with 3,3’,5,5’-tetramethylbenzidine (TMB) (Invitrogen, Cat# 229280010) for 3–5 minutes at RT, and the reaction quenched with H2SO4 1 M (Sigma-Aldrich, Cat# 4803641000). Plasma samples collected before the SARS-CoV-2 pandemic were also included as negative controls (same dilution of test samples), whereas an RBD-specific monoclonal antibody [ACROBiosystems Anti-SARS-CoV-2 Spike RBD Antibody, Chimeric mAb, Human IgG1 (AM130), Cat# S1N-M130] was included as positive control (diluted at 1:1, 1:6, 1:36). The optical density (OD) was measured by using PerkinElmer EnSight™ Multimode Plate Reader at 450 nm and 650 nm. Samples OD values were background-subtracted [(sample OD450 – sample OD650) – (blank OD450 – blank OD650)], and the area under the curve (AUC) was determined with GraphPad Prism 10.1.
RBD-blocking antibodies
RBD-blocking antibodies reactive against WT and BA.4/5 virus, which are antibodies inhibiting the binding of the spike RBD to its receptor ACE2 and thus a surrogate of neutralizing antibodies, were measured in EDTA plasma samples by commercially available kits [ACROBiosystems Anti-SARS-CoV-2 Neutralizing Antibody Titer Serologic Assay Kit (Spike RBD), Cat# RAS-N022, and ACROBiosystems Anti-SARS-CoV-2 (BA.4&BA.5) Neutralizing Antibody Titer Serologic Assay Kit (Spike RBD), Cat# RAS-N107], according to manufacturer’s instructions. Briefly, 50 μL/well of diluted plasma samples (1:100 in dilution buffer), positive control, and negative control were incubated with 50 μL/well of HRP-SARS-CoV-2 Spike RBD 1 μg/mL for 1 hour at +37 °C in the dark. After washing ×3, 100 μL/well of substrate solution were added and incubated for 20 minutes at +37 °C in the dark, followed by reaction quenching with 50 μL/well of stop solution. The optical density (OD) was measured by using PerkinElmer EnSight™ Multimode Plate Reader at 450 nm and 630 nm. Samples OD values were background-subtracted (OD450 – OD630), and results were expressed as percentage (%) of inhibition of RBD-ACE2 binding, calculated as: [1 − (sample OD/negative control OD)] × 100.
Statistics and reproducibility
Continuous variables were expressed as median (interquartile range, IQR), while categorical variables as number, n (percentage, %). Kruskal-Wallis or Friedman test with Dunn’s multiple comparison test, Wilcoxon test, and Mann-Whitney test were used for comparisons of continuous variables as appropriate. Fisher exact test was employed for comparison of categorical variables. Spearman’s correlation test was used to explore correlations between continuous variables. Multivariable linear regression analysis adjusted for potential confounders was used to explore: (i) the association between markers of HIV-related T cell dysfunction (CD4+CD127+ cells, CD8+CD38+ cells, CD4+CD45RA+ cells, CD8+CD45RA+ cells, and CD4/CD8 ratio; independent variables) and immune responses to the bivalent booster in PLWH (dependent variable); (ii) the association between HIV status (independent variable) and RBD-binding/blocking antibodies fold-change following the bivalent booster (dependent variable). Details on the statistical test used are reported in the respective figure legends. Data were analyzed and graphed with GraphPad Prism 10.1. Permutation test in SPICE 6.0 was employed to compare polyfunctionality patterns of SARS-CoV-2–specific Th1/Tc1 cells. P values less than 0.05 were considered statistically significant.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
