Adjuvants and formulations

The following adjuvants were used at the dilutions noted in the figure legends: SWE (squalene oil-in-water (OIW) emulsion, stock 4.1% squalene, Seppic, France), MPL (3D-6A PHAD, stock 1 mg/mL in DMSO, Avanti Polar Lipids, Alabaster, AL), QS-21 and QS-7 (stock 1 mg/mL in DMSO, Desert King, San Diego, CA), 3M-052 absorbed to 2 mg/mL Alum (AL030, stock 0.12 mg/mL, 3 M Corporate Research and Materials Laboratory (St. Paul, MN) and Access to Advanced Health Institute (AAHI, Seattle, Washington)), 3M-052 in 4% OIW (EM128, stock 0.12 mg/mL, 3 M/AAHI), and dmLT (stock 1 mg/mL in 700 μg of protein in 42.7 mM sodium phosphate, 10.7 mM potassium phosphate, 82 mM NaCl, 5% lactose, supplied by PATH (Seattle, Washington), manufactured by Walter Reed Army Institute of Research Pilot Bioproduction Facility (Silver Spring, MD)). LM, LQ, LMQ (liposomal (LS) formulations of MPL, QS-21 and MPL + QS-21 respectively), SM, SQ, and SMQ (OIW emulsions of MPL, QS-21 and MPL + QS-21 respectively) were obtained from the Vaccine Formulation Institute (VFI, Plan-les-Ouates, Switzerland). Non-MPL-containing stocks were confirmed to be free of endotoxin at the concentrations used (<1 EU/mL) using a limulus amoebocyte lysate (LAL) assay in accordance with the manufacturer’s protocol (Charles River, Wilmington, MA). Unformulated MPL, QS-21, and QS-7 were prepared in DMSO (Fisher Scientific, Hampton, NH), and DMSO controls were used as vehicle controls. Dynamic light scattering was performed on all formulations using a Zetasizer Nano ZS (Malvern Panalytical, Malvern, United Kingdom) to measure size, polydispersity index, and surface zeta potential over time to ensure formulation stability (Supplementary Fig. 1).

Blood collection

Peripheral blood samples from healthy study participants (n = 5; 26–45 years old) was collected after written informed consent was obtained in accordance with the Declaration of Helsinki and as approved by the Ethics Committee of Boston Children’s Hospital (BCH) (protocol number X07-05-0223). All participants were fully immunized against SARS-CoV-2 (>2 weeks prior) with one of the three FDA Emergency Use Authorized vaccines (Pfizer/BioNTech, Moderna, Janssen) with approval from the Ethics Committee of Boston Children’s Hospital, Boston, MA (protocol number X07-05-0223). Samples were de-identified, anti-coagulated with 20 U/mL pyrogen-free heparin sodium (Fresenius Kabi, Bad Homburg, Germany), and processed within 2–4 h.

Whole blood assay (WBA)

We adapted a previously described method to measure adjuvant activity in whole blood33 Briefly, whole blood was mixed 1:1 with sterile RPMI 1640 medium (Gibco, Ward Hill, MA, USA) and 180 µL of this was added to each well of a 96-well U-bottom plate containing 20 µL of freshly prepared adjuvant formulations at 10x the final concentration in RPMI 1640. The 200 µL/well suspensions were gently mixed by pipetting and were incubated at 37 °C for 24 h in a humidified incubator with 5% CO2. After incubation, the plate was centrifuged for 3 min at 500 × g. Supernatants were collected through pipetting without disturbing the cell pellet and were frozen at −80 °C until cytokine multiplex was run. Partial lysis of erythrocytes was obtained by incubating the cell pellet with BD FACS lysing solution (BD Biosciences, San Jose, CA, USA). Cells were washed three times with DPBS (Gibco, Ward Hill, MA, USA), fixated in 1% paraformaldehyde (Thermo Fisher Scientific, Ward Hill, MA, USA), and stored at 4 °C until flow cytometry analysis.

Isolation of mononuclear cells, monocytes and T cells, and generation of MoDCs

Mononuclear cells and monocytes were isolated as follows34. Heparinized blood from adult donors was centrifuged at 500 × g for 10 min. The upper layer of platelet-rich plasma was removed and further centrifuged at 3000 × g for 10 min. The platelet-poor plasma was stored at 4 °C for use in culture. Blood was restored to its original volume by adding DPBS and was carefully layered onto a Ficoll-Paque gradient (Cytiva, Marlborough, MA, USA). The mononuclear cell fraction was collected after centrifugation for 30 min at 1000 × g and was washed twice with DPBS. Monocytes were isolated from this cell fraction through positive selection with magnetic CD14 MicroBeads (Miltenyi Biotec, Auburn, CA, USA) in accordance with the manufacturer’s instructions. CD8+ and CD4+ T cells were then sequentially isolated from the remaining cell fraction in the same way, using CD8+ and CD4+ MicroBeads in accordance with the manufacturer’s instructions. CD14+ monocytes were cultured in a humidified incubator at 5% CO2 at 37 °C in 75 cm2 tissue culture flasks at a concentration of 106 cells/mL of RPMI 1640 supplemented with 1% penicillin-streptomycin (Gibco, Ward Hill, MA, USA), 10% FBS, 0.1% IL-4, and 0.1% GM-CSF (Miltenyi Biotec, Bergisch Gladbach, Germany) for five days. CD8+ and CD4+ T cells were frozen in 10% FBS (Gibco, Ward Hill, MA, USA), 10% DMSO and 80% RPMI, and were stored at −80 °C until use. After five days of culture, immature monocyte-derived dendritic cells (MoDCs) were harvested by removing the loosely adherent cell fraction through gentle pipetting.

MoDC assay

MoDCs were suspended in fresh MoDC culture medium at a concentration of 1.11 × 106 cells/mL, and a volume of 180 µL (200,000 cells) was plated into each well of a 96-well U-bottom plate containing 20 µL of freshly prepared adjuvant formulations at 10× the desired final concentration. Suspensions were mixed by gentle pipetting and the cells were then cultured for 24 h in a humidified incubator at 37 °C with 5% CO2. Cells were then centrifuged for 3 min at 500 × g, and supernatants were harvested and stored at −80 °C for use in further functional assays. Cells were washed with DPBS and fixated for flow cytometry.

MoDC:T cell interface assay (DTI)

DTI was performed as follows16. MoDCs generated as described above were suspended in RPMI 1640 supplemented with 1% penicillin-streptomycin (Gibco, Ward Hill, MA, USA), 10% FBS, 0.1% IL-4, and 0.1% GM-CSF. 25,000 cells were added per well in 90 µL to a 96-well U-bottom plate containing 10 µL of adjuvant formulations at 10× the desired final concentration in the presence or absence of recombinant SARS-CoV-2 Spike protein (Thermo Fisher Scientific, Ward Hill, MA, USA). Suspensions were mixed by gentle pipetting and cultured for 24 h. On the same day, CD4+ and CD8+ T cells were thawed and resuspended in RPMI 1640 with 10% FBS and 1% penicillin-streptomycin at a concentration of 3.33 × 106 cells/mL. T cells were cultured for 24 h in a six-well plate with 3 mL per well, and then harvested by gentle pipetting, resuspended at a concentration of 2.5 × 106 cells/mL. 100 µL of either CD4+ or CD8+ T cells was added to each well of the respective MoDC plate, and the co-culture was incubated for four days. Cells were then centrifuged for 3 min at 500 × g. Supernatants were harvested and stored at −80 °C for use in further functional assays. Cells were washed with DPBS, fixated in 1% PFA, and stored at 4 °C for flow cytometry.

Human tissue construct (HTC)

A human tissue construct model of monocyte extravasation and in-tissue autonomous differentiation was generated as we have previously described13. Briefly, Endotoxin-free human type I collagen Advanced BiomatrixTM (San Diego, CA) cushions were cast in 96-well microtiter plates (Costar round bottom, Thermo Fisher Scientific Inc., Ward Hill, MA, USA). Human type I collagen cushion solution was prepared by mixing 10x M199 media (Gibco, Ward Hill, MA, USA), 0.1 N NaOH, and human collagen (3 mg/mL) at a proportion of 1:5:8 respectively. 70 µL of this solution was applied to each of the inner 60 wells of a 96-well flat-bottom plate using a repeat pipette, and 200 µL of HBSS (Gibco, Ward Hill, MA, USA) was added to the 36 outer wells to act as an evaporation barrier. The plate was incubated at 5% CO2/37 °C for 24 h to congeal. Meanwhile, using Trypsin-EDTA and the same M199 media containing 50% FBS and 1% Penicilin/Streptomycin/Glutamine (Gibco, Ward Hill, MA, USA), 85–90% confluent human umbilical vein endothelial cells (HUVEC) cultures (as assessed with an inverted microscope) were passed to larger (150 cm2) vented cap tissue culture flasks pre-coated with human fibronectin (0.5 mg/mL, Biomedical Technologies Inc. Stoughton, MA) and incubated at 5% CO2/37 °C. Ten minutes before seeding endothelial cells, collagen cushions were prepared by adding 100 µL of M199 with 1% Penicilin/Streptomycin/Glutamine (Gibco, Ward Hill, MA, USA) for 10 minutes before aspirating out and adding 20 µL of 0.5 mg/mL human fibronectin. Typically, one T150 flask of confluent HUVECs contained sufficient cells to coat two cushion plates (120 wells). Prior to seeding, excess fibronectin was aspirated out and cells from one confluent T150 flask were resuspended in approximately 13 mL of media to be dispensed as 100 µL per well (120 wells). Plates were incubated at 5% CO2/37 °C until confluency of monolayer. Monolayer integrity, confluence, and morphology of endothelial cells were assessed by inverted microscopy using phase contrast at 4× magnification (Nikon TS100 inverted microscope EVOS XL Core Imaging System (Fisher Scientific, Hampton, NH, USA)). Only 100%-confluent TCs were used for testing. CD33+ cells from each donor were plated on top of autologous platelet-poor plasma. The adjuvant formulations were immediately added to the appropriate wells from a pre-prepared dilution plate. The plates were incubated for 48 h at 37 °C in 5% CO2. After incubation, representative images were taken, and supernatants were collected without disturbing the endothelial layer. The reverse-transmigrated cells were then resuspended in DPBS and transferred to a fresh 96-well U-bottom plate. 10 µL of sample was removed for counting in a hemocytometer after dilution in Trypan Blue (Gibco, Ward Hill, MA, USA). Remaining cells were washed again in DPBS, fixated in 1% PFA (Fisher Scientific, Hampton, NH, USA), and stored at 4 °C for flow cytometry.

Cytokine multiplex

Cytokine profiles of supernatants derived from whole blood assay, MoDC assay, DTI assay, and HTC assay were analyzed using a multianalyte fluorescent bead-based array (Luminex Corp., Austin, TX, USA). Quantification of cytokines was done using a custom built 21-plex kit using the Milliplex HCYTA-60K Human Cytokine/Chemokine/Growth Factor Panel A (Millipore, Merk, Darmstadt, Germany). This customized kit included CXCL8, CXCL10, GRO, IFNa-2, IFNγ, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-10, IL-12p40, IL-12p70, IL-13, IL-17, IL-18, IL-27, MCP-1, MIP-1a, and TNF, and was run according to the manufacturer’s instructions. Sample fluorescence data was collected using a Flexmap 3D analyzer running xPONENT software version 4.2, and results were fit to a 5-point log curve and converted into pg/mL values using manufacturer-provided standard solutions and Milliplex Analyst software version 5.1.

Flow cytometry

Flow cytometry was used to identify and characterize cell subpopulations following whole blood assay, MoDC assay, DTI assay, and HTC assay. Flow cytometry was performed using an LSRFortessa flow cytometer (Becton Dickinson) and analyzed with FlowJo software version 10. Cells were stained for 30 min at 4 °C in the dark with panels which used the following antibodies: anti-CD14-PE (Clone M5E2, catalog (cat)# 557154), anti-CD56-PerCP-Cy5.5 (Clone B159, cat# 560842), anti-CD20-PE-Cy7 (Clone L27, cat# 335793), anti-CD86-AF700 (Clone 2331, cat# 561124), anti-CD197-PE-Cy7 (Clone 3D12, cat# 557648), anti-CD4-V450 (Clone RPA-T4, cat# 560345), anti-CD8-V450 (Clone RPA-T8, cat# 561426), anti-CD134-PE (Clone L106, cat# 340420), anti-CD154-APC (Clone 89-76, cat# 648887), anti-CD25-FITC (Clone M-A251, cat# 555431), anti-CD16-FITC (Clone B73.1, cat# 561308), anti-CD40-FITC (Clone 5C3, cat# 555588) (BD Biosciences, East Rutherford, NJ), anti-CD3e-AF647 (Clone UCHT1, cat# A51001) (Invitrogen, Waltham, MA), anti-CD123-APC-eFluor780 (Clone 6H6, cat# 47-1239-42) (eBioscience, San Diego, CA), anti-CD1c-PB (Clone L161, cat# 331508), and anti-HLA-DR-BV605 (Clone L243, cat# 307639) (Biolegend, San Diego, CA). All antibodies were added at 1:25 (v/v) dilution. Cells were then washed and resuspended in DPBS prior to data acquisition. All representative gating strategies are shown in Supplementary Fig. 3.

T-SNE analysis in Fig. 2 was run using Flowjo, All flow cytometry samples from all treatments combined were used to map t-SNE populations. Uncompensated parameters were selected, additional parameters were set to 1000 iterations, complexity 30, and a learning rate of 7% total events mapped. Vantage point tree k-nearest neighbors (KNN) algorithm was used in combination with Barnes-Hut gradient algorithm.

Mouse immunization

All animal studies complied with all relevant ethical regulations for animal testing and research set by the Swiss Federal Law on the Protection of the Animals and the Association for the Assessment and Accreditation of Laboratory Animal Care International (AAALAC) and have received ethical approval from the animal care and use committee (IACUC) regulatory committee at Boston Children’s Hospital (protocol number 00001573). Animals were co-housed with food (isopro RMH3000 irradiated) and water ad libitum. After completion of all experiments, animals were euthanized following institutional guidelines, specifically by regulated exposure to CO2, with confirmation by paw-pinch and cervical dislocation. For evaluation of surrogate virus neutralization titer (sVNT), six- to eight-week-old female C57BL/6J mice (Charles River) were immunized intramuscularly (hind leg) on days 0 and 21 with 50 µL of adjuvanted SARS-CoV-2 Spike antigen formulations, or controls including excipient and SARS-CoV-2 Spike antigen alone. Blood samples were collected on days 20 and 42, incubated at RT for 2 h to allow clot formation and centrifuged at 1000 × g for 10 min. Serum layers were then collected and stored at −80 °C prior to analysis. SARS-CoV-2 spike receptor binding domain antibodies capable of blocking human ACE2 receptor binding were detected in sera using sVNT, (Genscript, Cat# L00847) specific for wild-type RBD (Z03594) according to the manufacturer’s instructions (Genscript,). Briefly, sera and control samples were prediluted before being mixed 1:1 with an HRP-RBD solution. After 30 minutes incubation at 37 °C, 100 µL of this preparation was added to the capture plate pre-coated with ACE2 human cell receptor and incubated for 15 min at 37 °C. After four washes, a TMB solution was added and the plate was incubated for 15 min at RT. The reaction was quenched with 1 M sulfuric acid and absorption was read at 450 nm using a microplate reader.

A control cohort of animals was immunized with institutional animal care and use committee (IACUC) regulatory approval at Boston Children’s Hospital. Adjuvant formulations as indicated in Supplementary Fig. 4 were co-formulated with SARS-CoV-2 Spike antigen, for analysis of anti-SARS-CoV-2 Spike antibodies by ELISA following immunization. Six- to eight-week-old BALB/cJ mice (Jackson Laboratories), were intramuscularly immunized (IM) in the caudal thigh with adjuvanted formulations as indicated, containing 1 µg of recombinant wild-type Spike protein (Wuhan human-1 isolate, GenBank MN90894, M1-Q1208), produced in HEK293 cells35. Mice were immunized following a prime (day 0), boost (day 14) immunization schedule. At 24 h post-prime and post-boost immunization, blood was collected in heparinized capillary tubes by retro-orbital (RO) bleed, stored on ice, and plasma isolated by centrifugation (500 × g for 10 min, room temperature) to evaluate the degree of reactogenicity-associated analyte induction. IL-6, TNF, and IL-1β were quantified by ELISA following the manufacturer’s recommendations (Invitrogen, 88-7064 (IL-6), 88–7324 (TNF), 88–7013 (IL-1β)). At 28 and 42 days post-prime immunization non-heparinized capillary tubes were used to collect blood by RO bleed and subsequently collect serum was collected by centrifugation within 2 h (1500 × g, 7.5 min). Serum was then transferred to a new microcentrifuge tube, and centrifuged again to maximize separation from clot, and increase serum volume collection. Anti-spike antibody quantification was performed as described previously35. Briefly, high-binding flat 96-well plates were coated overnight with 25 ng Spike antigen per well. After washing with 0.05% Tween20 in PBS plates were blocked with 1% BSA in PBS for 1 h, with incubation at room temperature (RT). Following an initial dilution of 1:100, 10 serial fourfold dilutions were applied to spike-coated plates for 2 h (RT). Following three washes, a 1 h incubation with horse radish peroxidase anti-mouse IgG, IgG2a, and IgG1, and another five washes, wells were incubated 5 min with tetramethylbenzidine (TMB; BD biosciences OptEIA Substrate Solution), then were inactivated with 2 N H2SO4. 450 nm absorbances were compared to blank, and interpolated titers were calculated based on the threshold crossing three times per-plate median blank values, and non-responsive samples were assigned a value of half lower limit of detection, 50, for statistical comparison.

Statistical analyses

Statistical analyses were performed using GraphPad Prism version 9.3.1 and R (version 4.2.1) with RStudio (version 2023.03.1 + 446). Dose-response curves were compared using 2-way repeated measures ANOVAs, and multiple comparison testing was performed using either Dunnet’s Method when comparing against the vehicle only or Tukey’s method when making comparisons across additional groups. A normal two-way ANOVA with Tukey’s correction for multiple testing was used to analyze the HTC results as well as sVNT and antibody quantification results, as comparisons were only made at a single concentration. Synergy was calculated using an adaptation of the Loewe method of additivity34. D values greater than 1 were considered antagonistic, D values equal to 1 were considered additive, and D values less than 1 were considered synergistic. Humoral immunity from mice was evaluated for normality by Shapiro-Wilk followed by Kruskal-Wallis and two-sided Wilcoxon tests. Flower plots were generated with Grapher v15.2.311. Graphical illustrations (Figure panels 1A, 2A, 3A, 4A) were generated with Biorender.com under publication agreement JV25VQMXXO.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.