Ethics statement

All animal experiments were conducted according to the animal study proposal (ASP LIR21/LIRID9) approved by National Institutes of Health (NIH), National Institute of Allergy and Infectious Diseases (NIAID), Animal Care and Use Committee (ACUC) that meets all federal requirements, as defined in the Animal Welfare Act (AWA), the Public Health Service Policy (PHS), and the Humane Care and Use of Laboratory Animals in AALAC accredited facilities. This study used 6–8-week-old female NZW rabbits (Charles Rivers Laboratories, Wilmington, MA, USA) that were co-housed, all of which were processed for terminal bleed collections under general anesthesia and euthanized by exsanguination as approved by the AVMA (American Veterinary Medical Association) and adopted by NIH-NIAID ACUC.

Construct design

The Env sequence for VLP and adenovirus constructs was derived from the clade C 1086 isolate (name: 1086-B2 C, GenBank accession number: FJ444395, from Malawi 2004)¹⁶. The original 1086 C was modified to introduce a K160N mutation to permit binding of antibodies specific for the apex (HxB2 numbering). The SOSIP version was generated by adding the following changes: a TPA signal sequence (MDAMKRGLCCVLLLCGAVFVSPSQEIHARFRRGAR), A501C and T605C (gp120-gp41_ECTO disulfide bond), I559P in gp41_ECTO (trimer-stabilizing), H66R and T316W (trimer-stabilizing), Q543N in gp41_ECTO (improved trimerization) and REKR to RRRRRR (R6) in gp120 (furin cleavage enhancement)^2,50,51 and a stop codon after residue 704. A second intermolecular disulfide bond was added by introducing mutations A73C in gp120 and A561C in gp41⁵². Eight BG505 Trimer Derived (TD) mutations were also introduced (E47D, K49E, V65K, E106T, E429R, R432Q and E500R; the sequence already contains a L at position 165)^10,53 and MD39 mutations (T106E, R304V, A319Y, P363Q, F519S, L568D, V570H and R585H; the sequence already contains an I at position 271 and a L at position 288)⁵⁴ resulted in 1086c SOSIP.v8.2 gp145.

To generate an NFL trimer design, the furin cleavage site REKR (HIV-1 Env residues 508–511) was replaced by a flexible linker (GGGGSGGGGS) to covalently link the gp120 and gp41 Env subunits²⁰. The natural HIV-1 Env leader sequence was replaced by the CD5 leader sequence to increase expression. The following HIV-1 Env TD substitutions were made to generate highly stable and homogeneous NFL trimers: E47D, K49E, V65K, E106T, E429R, R432Q, E500R; helix-destabilizing gp41 mutations, I559P, L568G, N636G; V3 and Fusion peptide stabilizing mutations, N302Y, T320M, F519R, L520R and V513Y. To further enhance sensitivity to the V2-apex antibodies, the K166R and H170Q mutations were also introduced^3,10. Finally, a second linker GGGGS was incorporated to replace the MPER, residues 664–683, to covalently link the 1086c NFL to the NDV TM. The Ad4 FDE3 Env150 was constructed as previously described except that a stop codon was introduced at position 732 to enhance surface expression, and the REKR furin cleavage site was restored to improve antigenicity⁵⁵. The 1086c Env also included the K160N mutation.

Plasmids and VLP production

Plasmids for the production of NDV VLP were generated as previously described¹⁴. Briefly, B1 strain of NDV cDNA sequences encoding NP, M and HN were subcloned into the mammalian expression vector pCAGGS to generate plasmids for co-transfection in the construction of VLPs. Chimeric constructs were constructed by combining protein coding regions of either Influenza H5 HA, HIV-1 Env 1086 C, HIV-1 Env 1086 C SOSIP, HIV-1 Env 1086 C NFL-TD or SARS-CoV2 S Delta+ with the TM and CT from NDV fusion (F) protein. F/H5, F/1086 C, F/1086 C SOSIP, F/1086 C NFL-TD and F/CoV2 Delta+ plasmids were generated by synthesizing the chimeric codon optimized F protein sequence containing ectodomain sequence from Influenza H5 HA (A/Vietnam/1203/2004; GenBank accession EF541402), HIV-1 1086 C Env, SOSIP stabilized HIV-1 1086 C Env, or NFL-TD stabilized HIV-1 1086 C Env respectively (Genscript, Piscataway, NJ, USA). The SARS-CoV-2 Delta+ full-length spike protein sequence (GenBank accession number OK098887) included additional mutations from other circulating Delta strains (R21T, T77K, E154K, Q216H, E482Q and H1099D) and retained the TM from SARS-CoV2 Spike but maintained the NDV/F CT, which was truncated to increase surface Spike. All chimeric constructs were subcloned into pCAGGS mammalian expression vector. RNA40D6 plasmid was generated by synthesizing the DNA fragment containing RNA40 sequence²⁸ between NDV leader and trailer sequences⁵⁶, and subcloned into the pCAGGS mammalian expression vector. An extra 3 nucleotides downstream of the NDV leader sequence were included to adjust the total number of NDV/RNA40 nucleotides divisible by 6.

VLPs were produced as previously described with some modifications¹⁴. The transfection was performed using the Expi 293 Expression System Kit [A14635] (Thermo Fisher Scientific, Waltham, MA, USA). Expi293F [A14527] (Thermo Fisher Scientific) cells were transfected using Expifectamine transfection reagent (Thermo Fisher Scientific) as recommended by the manufacturer. Briefly, 60 μg of plasmid mixture (For VLP H5: NP, M, HN and F/H5-vtn at the molar ratio of 1:1:1:1. For VLP-1086C-SOSIP or NFL-TD: NP, M, F/1086 C SOSIP or NLF-TD at the molar ratio of 1:1:1. For VLP CoV2: NP, M and F/CoV2 Delta+ at the molar ratio of 1:1:1 with 0.5 µg of TMPRSS2 plasmid) was transfected into 1.5 × 108 Expi 293 F cells. For all VLPs with TLR agonist incorporated, RNA40D6 plasmid was included in each plasmid panel at the same molar ratio. Expifectamine 293 transfection enhancers and 10 µg/mL of heparin were added to the culture 24 h post-transfection. Media containing VLPs were collected at 48 and 72 h post-transfection (72 and 96 h for VLP SOSIP or VLP NFL-TD) and purified by a series of discontinuous sucrose gradients, as previously described¹⁴. The media was centrifuged at 38,500 × g for 18 h at 4 °C using a SW32 rotor in an Optima L-100K Ultracentrifuge (Beckman Coulter, Brea, CA, USA) to obtain a VLP pellet. The VLP pellet was then resuspended in TNE buffer (25 mM Tris-HCl, 150 mM NaCl and 5 mM EDTA, pH 7.4) and layered on top of a discontinuous sucrose gradient containing 20% and 65% sucrose (w/v) cushion, followed by centrifugation at 100,000 × g for 6 h at 4 °C in a SW41 Ti rotor (Beckman Coulter). The VLP fraction collected between the 20% and 60% sucrose interphase was adjusted to 60% sucrose concentration and layered between a 50% and 80% sucrose gradient. The tube was then topped up with 10% sucrose and centrifuged at 200,000 × g for 16 h at 4 °C. The interphase between 10% and 50% sucrose containing purified VLPs was collected, diluted in TNE buffer, and pelleted by centrifugation at 145,000 × g for 6 h at 4 °C. The VLP pellet was resuspended in TNE buffer and stored at −20 °C until further use.

Western blot

Cell lysates or purified VLPs were heat-denatured at 95 °C for 10 min in sample buffer under reducing conditions. Samples were resolved on 10% Tris-Glycine SDS-PAGE and transferred on a nitrocellulose membrane for Western blot analysis using the following antibodies: chicken anti-Newcastle Disease Virus polyclonal antibody [ab34402] (Abcam, Cambridge, United Kingdom) and goat anti-chicken IgY H&L- HRP conjugated secondary antibody [ab6877] (Abcam) for detection of Newcastle Disease Virus antigen; rabbit anti-HIV-1 gp120 Env (Clade B, IIB) antibody [ABL#5414] (Advanced Bioscience Laboratories, Rockville, MD, USA) and donkey anti-rabbit IgG (H + L) cross-absorbed- HRP conjugated antibody [SA1-200] (Thermo Fisher Scientific) for detection of HIV-1 Env protein; mouse anti-influenza A virus (H5N1/HA1) antibody [ab135382] (Abcam) and horse anti-mouse IgG-HRP conjugated antibody [7076S] (Cell Signaling Technology, Danvers, MA, USA) for detection of influenza HA protein; rabbit anti-spike (SARS-CoV2) antibody [scv2-SA-200] (eEnzyme LLC, Gaithersburg, MD, USA) and donkey anti-rabbit IgG (H + L) cross-absorbed-HRP conjugated antibody [SA1-200] (Thermo Fisher Scientific) for detection of SARS-CoV-2 spike proteins. Signals were detected using the SuperSignal West Pico Plus Chemilluminescent substrate [34580] (Thermo Fisher Scientific) with the ChemiDoc MP Imaging system (Bio-Rad Laboratories) and analyzed by ImageJ (Version 2.1.0/1.53c).

Flow cytometry

To examine the preservation of native-like Env conformation, HIV-1/NDV chimeric proteins were expressed in mammalian cell lines, and the binding of anti-Env antibodies was measured by flow cytometry. One day before transfection, 4.5 × 10⁶ A549 human adenocarcinoma cells [CCL-185] (ATCC, Manassas, VA, USA) were seeded in T-75 flasks with F-12K medium [30–2004] (ATCC) containing 1% Penicillin-Streptomycin-Glutamine [10378016] (Thermo Fisher Scientific) and 10% Fetal Bovine Serum [100–106] (Gemini Bio-Products, West Sacramento, CA, USA). Cells were transfected with 15 µg of NDV-F/Env chimeric plasmid DNA, 75 µl of DNA-In® A549 Transfection Reagent [73772] (MTI-Global Stem, Gaithersburg, MD, USA), and 150 µl of Opti-MEM [31985070] (Thermo Fisher Scientific) and cultured for 48 h at 37 °C with 5% CO₂. To detect expression of Env, cells were collected 48 h post-transfection with 0.01 M EDTA in phoshpate-buffered saline (PBS) and stained with 50 µl of anti-Env monoclonal antibodies PGT145, PG16, VRC01, b12, PGT151, 8ANC195, 35O22, 10E8, F105, or 447-52D (BEI Resources, Manassas, VA, USA) at 1 µg/ml in PBS containing 0.01 M HEPES and 0.09% bovine serum albumin [A7979] (Millipore Sigma, Burlington, MA, USA) for 1 hr at 37 °C. A secondary antibody, goat anti-human IgG Fab2-phycoerythrin (PE) [109-116-097] (Jackson ImmunoResearch, West Grove, PA, USA) was used at a 1:100 dilution for 1 hr at 37 °C. To differentiate live and dead cells, a Live/Dead Fixable Violet Dead Cell Stain Kit [L34964] (Thermo Fisher Scientific) was used at a 1:250 dilution for 30 min at room temperature. Cells were fixed with 250 µl of Cytofix/Cytoperm [554722] (Becton Dickinson, Franklin Lakes, NJ, USA) for 20 min on ice. Alternatively, A549 cells were infected with Ad4-Env at an MOI of 0.1, harvested at 48 h post-infection, and processed for surface staining. Cells were then permeabilized overnight in Perm/Wash buffer [554723] (Becton Dickinson) and intracellularly stained with 50 µl of anti-Hexon (adenoviral capsid protein) antibody 8C4-allophycocyanin (APC) [NB600-413APC] (Novus Biologicals, Centennial, CO, USA) at a 1:700 dilution in Perm/Wash buffer for 30 min on ice. Cells were analyzed by flow cytometry on a BD FACS Aria with FACSDiva software (Becton Dickinson).

To assess incorporation of viral glycoproteins, anti-Env antibodies, VRC01 or PGT145 were custom conjugated to PE (Becton Dickinson). VLPs were mixed with a diluted fluorescent primary anti-Env antibody in PBS containing bovine serum albumin (BSA) in 4.5-ml V-bottom polystyrene tubes and incubated for 30 min in the dark at 4 °C. After incubation, the sample was diluted 10× in PBS/BSA. VLPs were transferred to 5-ml polystyrene round-bottom tubes and were analyzed with a FACSymphony S6 cell sorter with FACSDiva software (version 10.9.0) (Becton Dickinson). VLPs not containing HIV-1 Env were stained with the fluorescent anti-Env antibodies to use as controls. The cytometer was set to trigger on both forward scattering (FSC) and side scattering (SSC) lights. VLPs were detected by FSC and SSC, and then the population of gated virions was determined to be expressing Env using fluorescence emitted from the anti-Env PE conjugated antibodies. To confirm that events were indeed VLPs, the FSC and SSC thresholds and voltages were adjusted to discriminate buffer particulates from VLPs using PBS/BSA without VLPs. Cleaning with BD Detergent Solution [660585] (Becton Dickinson) was performed as needed between each sample to ensure fewer than 50 events were detected in a tube of PBS/BSA collected over a minute at the maximum flow rate.

Quantitative PCR (qPCR)

qPCR was performed using the QuantStudio 3 System (Thermo Fisher Scientific) to measure the levels of RNA40 incorporated into VLP variants. RNA was extracted from VLPs by QIAamp Viral RNA Mini Kit [52904] (Qiagen, Venlo, Netherlands) with and without RNase treatment, then reverse transcribed by SuperScript III First-Strand Synthesis System [18080051] (Thermo Fisher Scientific) using random hexamers according to the manufacturers’ instruction. Synthesized cDNA product was then combined with TaqMan Fast Advanced Master Mix [4444556] (Thermo Fisher Scientific). The RNA40D6 transcript was amplified using the following synthesized primers: forward primer 5′-CCAAAGAGTCGGAATTTAACGC-3′, reverse primer 5′-TGTGAGGTACGATAAAAGGCG-3′, and TaqMan probe labeled with a 5′ reporter dye (FAM) and 3′ fluorescent quencher (TAMRA dye): 5′ (6-FAM)-ACGGAGTCACACAACAGACGGG- (TAMRA-Sp) 3′. The reaction conditions were as follows: one 20 s period at 95 °C, followed by 40 cycles of 1 s at 95 °C and 20 s at 60 °C. The Cq values were used to report the level of transcripts detected in copies/µg. Data were analyzed using the QuantStudio 3/5 Real-Time PCR software and Thermo Fisher Connect Platform (Thermo Fisher Scientific).

Negative-stain electron microscopy

A 4.8-µl drop of the sample was applied to a freshly glow-discharged carbon-coated copper grid for ~15 s and removed using blotting paper. The grid was washed with three drops of buffer containing 0.01 M HEPES, 150 mM NaCl, pH 7.0, followed by negative staining with three drops of 0.75% uranyl formate. Micrographs were acquired using a Talos F200C transmission electron microscope (Thermo Fisher Scientific) operated at 200 kV and equipped with a Ceta CCD camera (Thermo Fisher Scientific). The nominal magnification was 57,000, corresponding to a pixel size of 2.53 Å. To estimate the number of visible spikes per VLP, micrographs were high-pass filtered to 250 Å using SPIDER⁵⁷ to suppress the signal corresponding to the VLP, followed by low-pass filtration to 15 Å to eliminate high-frequency noise. Spikes were then detected automatically in Relion 3.0⁵⁸ using a Laplacian-Gaussian filter with a minimal diameter of 90 Å and a maximal diameter of 150 Å. When the elongated shape of the spikes and their high density prevented reliable automatic quantification, visible spikes were counted manually instead.

Rabbit immunization

Female NZW rabbits [Crl: KBL(NZW), stock number 052] (6–8 weeks old: 6 animals per group) (Charles River Laboratories) were immunized intramuscularly with 150–500 µg of purified VLPs in TNE buffer at 0, 4, 12, and 20 weeks and 10¹¹ TCID₅₀ of Ad4 FED3 Env150 at 0 and 4 weeks for bolus immunization studies. One-fifth of a human dose of AS01 adjuvant (containing 10 µg of 3-O-desacyl-4’monophosphoryl lipid A (MPL) from Salmonella minnesota and 10 µg of QS-21, a saponin, combined in a liposomal formulation (GlaxoSmithKline Biologicals, London, United Kingdom) was formulated with VLPs for the groups tested to assess the impact of the adjuvant. For dose escalation immunization studies, a total of 500 µg of VLP was split into 7 doses (1, 2, 5.8, 15.8, 42.9, 116.5 and 316 µg) and administered intramuscularly in 48-h intervals. The adjuvant was also split proportionally to the VLP in these groups. The blood samples were collected at 0, 4, 8, 12, 14, 16, 20, 22, 24, and 28 weeks post-immunization.

Purification of serum Ig

Serum Ig was purified using a 1:1 mix of rProtein A Sepharose™ Fast Flow and Protein G Sepharose™ 4 Fast Flow resins [17-1279-02 and 17061802] (Cytiva, Marlborough MA, USA) according to manufacturer’s instructions. Briefly, 1 ml of rabbit serum from each animal was diluted (1:1) in Pierce™ Protein A/G IgG binding buffer [54200] (Thermo Fisher Scientific) and passed through a Poly-Prep® Chromatography gravity flow column [7311550] (Bio-Rad Laboratories, Hercules, CA, USA) packed with 500 µl of Protein A/G Sepharose. Columns were washed with PBS to remove non-specific binding and eluted with IgG Elution Buffer (Thermo Fisher Scientific). Purified Ig was dialyzed in PBS and concentrated by centrifugation at 3000 × g at 4 °C using a 30 kDa MWCO Amicon ultra centrifugal filter [UFC903008] (Millipore Sigma). Pierce™ BCA Protein Assay Kit [A55864] (Thermo Fisher Scientific) was used to quantify concentrated purified Ig and stored at −80 °C until use.

Mesoscale discovery assay

HIV-1 Env-specific antibodies were assayed using the 384-well Streptavidin SECTOR Plate [L21SA-1] (Meso Scale Discovery, Rockville, MD, USA). To reduce non-specific binding signals, plates were blocked using a 5% MSD Blocker A solution [R93BAA] (Meso Scale Discovery) for 1 h with shaking using an Orbi-Shaker MP (Benchmark Scientific Inc., Edison, NJ, USA) at room temperature. Following blocking, plates were washed with 1× MSD Wash Buffer [R61TX-1] (Meso Scale Discovery) and incubated with the biotinylated capture protein with shaking for 1 h at room temperature. Plates were washed with 1× MSD Wash Buffer, and then serial dilutions of samples and controls were prepared in 1% MSD Blocker A [R93BA] (Meso Scale Discovery) in DPBS with 0.05% Tween-20, and then added to the plate and incubated for 1 h with shaking at room temperature. Samples were tested using serial dilutions starting at a minimum dilution of 1:100. After sample incubation, plates were washed with 1× MSD Wash Buffer, and 1 μg/mL goat anti-rabbit SULFO-TAG™ conjugated detection antibody [R32AB-1] (Meso Scale Discovery) was added for 1 h with shaking at room temperature. To detect signals 1× MSD Read Buffer was applied and analyzed using the MSD Sector Imager S600 (Meso Scale Discovery). All samples were tested in duplicate. Samples with a replicate coefficient of variation >30% were retested. Serial dilutions of the sample were used to assign an area under the curve (AUC) value as the primary readout. Results were plotted and analyzed using Prism version 9 or newer (GraphPad, San Diego, CA).

Electron microscopy-based polyclonal epitope mapping

The details of serum and sample preparation to obtain polyclonal antigen-binding fragments (Fabs) for EMPEM were previously described⁵⁹. Briefly, IgG was isolated from 1 mL rabbit sera (drawn at week 31 post-first immunization) using a self-packed FPLC column containing 5 mL CaptureSelect Fc multispecies resin [2942852005] (Thermo Fisher Scientific) on an AKTA Pure system (Cytiva). Polyclonal IgG was eluted with 0.1 M glycine pH 2.0, and buffer exchanged into TBS (50 mM Tris-HCl, 150 mM NaCl, pH 7.4). Papain [76216] (Millipore Sigma) was used to digest IgG to Fabs. Trimer-Fab complexes were prepared and incubated overnight by mixing 15 µg of 1086 C NFL, BG505 SOSIP or PVO.04 NFL trimers with 1 mg of Fab mixture (containing Fc and residual papain). The complexes were purified using a Superdex 200 Increase 10/300 GL gel filtration column [28990944] (Cytiva). Purified complexes were concentrated and diluted to a final concentration of 0.03 mg/mL, which were adsorbed on glow-discharged carbon-coated copper mesh grids and stained with 2% (w/v) uranyl formate. Electron microscopy images were collected on an FEI Tecnai Spirit T12 equipped with an FEI Eagle 4k × 4k CCD camera (120 keV, 2.06 Å/pixel) or a FEI Thermo Fisher Scientific Glacios equipped with a Thermo Fisher Scientific Falcon IV direct electron detector (200 keV, 1.89 Å/pixel) and processed using Relion 3.0⁶⁰ following standard 2D and 3D classification procedures. UCSF Chimera⁶¹ was used to generate the composite maps and estimate epitope contacts by fitting the atomic coordinates of a BG505 SOSIP protomer into the map.

Pseudovirus production and entry inhibition assay

HIV-1 Env, Flu-HA and SARS-CoV-2 S pseudoviruses were generated as previously described⁴⁴. To produce HIV-1 Env pseudovirus, HEK293T cells [CRL-3216] (ATCC) were co-transfected with plasmids encoding an Env-deficient backbone (pSG3ΔEnv) and HIV-1 Env (BG505, Q769.d22, KER 2018, SF162, BaL.01, PVO.04, Du156.12, ZM106.9, 16055, 1086 C) or SIV Env (SIVmac256) at a ratio of 3:1. For influenza virus HA pseudovirus, HEK293T cells (ATCC) were transfected with the following plasmids: 1 µg of HA, 0.1 µg of matched NA (A/Vietnam/1203/2004), 17 µg of pCMV-Luc and 0.1 µg of TMPRSS2. SARS-CoV2 pseudoviruses were generated by transfection of HEK293T/17 cells [CRL-11268] (ATCC) with the following plasmids: 0.53 µg of SARS-CoV-2 S (Wuhan, Delta+ S, Omicron B.1, or Omicron BA.2), 9.2 µg of lentiviral backbone (VRC5602), 9.2 µg of pCMV-Luc and 0.16 µg of TMPRSS2. Supernatants containing pseudoviruses were harvested 48-72 h post-transfection and processed by centrifugation at 2000 × g for 10 min at room temperature and filtration through a 0.2 µm filter for titration. Processed pseudoviruses were titered using the following cell lines: TZM-bl [HRP-8129] (BEI Resources) for HIV-1, HEK293A [R70507] (Thermo Fisher Scientific) for Influenza, and HEK293T/ACE [631289] (Takara Bio Inc., Kusatsu, Shiga, Japan) for SARS-CoV2 and stored at −80 °C until further use.

HIV-1, Influenza and SARS-CoV-2 neutralization activities of sera from VLP-HIV-1/Flu/CoV2 immunized rabbits were tested using a single-round pseudovirus infection of TZM-bl, HEK293A or HEK293T/ACE cells, respectively, as described previously²³. For testing HIV-1 and Influenza neutralization, heat-inactivated rabbit serum or purified serum Ig was serially diluted five-fold with Dulbecco’s modified Eagle medium (Thermo Fisher Scientific) supplemented with 10% FCS (GeminiBio), and 10 μl of serum, purified Ig, or mAb was incubated with 40 μl of pseudovirus in a 96-well plate at 37 °C for 30 min. TZM-bl (for HIV-1), HEK293A (for Influenza) or HEK293T/17-Ace (for SARS-CoV-2) cells at 1 × 10⁴ per well were then added, and plates were incubated at 37 °C with 5% CO₂ for 48 h (for HIV-1 and Influenza) or 72 h (for SARS-CoV-2). Luciferin signals were then detected using a Luciferase Assay System (Promega), and the relative light units (RLU) were read on a Victor X2 luminometer (Perkin Elmer, Waltham, MA, USA). All neutralization assays were performed in duplicate, and means are reported. All neutralization assays were repeated at least once with similar results.

To competitively inhibit V3-mediated SF162 pseudovirus neutralization, 10 μl of serially diluted heat-inactivated rabbit serum was co-incubated with 20 μl of solubilized V3 peptide or control peptide (EIYKRWII) at a concentration of 75 μg/μl for 90 min at 37 °C. Pseudovirus (20 μl) was then added before incubation at 37 °C for 30 min. TZM-bl cells (1 × 10⁴ cells/well) were then added to 96-well cell culture plates and incubated, as described above.

Statistical methods

Welch’s two-sample t test (unequal variances) was used for all comparisons unless otherwise noted. ArUC was calculated by multiplying neutralizing magnitude at each timepoint by the number of timepoints using trapezoidal integration using R ve4.4.0. In cases where experiment length differed, the duration was shortened to match for comparison. To assess V3-mediated neutralization a paired t test was used to compare control peptide to V3 peptide.

Reporting summary

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