henhaobio
Sign in

Abstract

Recruitment and expansion of rare precursor B cells in germinal centers (GCs) is a central goal of vaccination to generate broadly neutralizing antibodies (bnAbs) against challenging pathogens such as HIV. Multivalent immunogen display is a well-established method to enhance vaccine-induced B cell responses, typically accomplished by using natural or engineered protein scaffolds. However, these scaffolds themselves are targets of antibody responses, with the potential to generate competitor scaffold-specific B cells that could theoretically limit expansion and maturation of "on-target" B cells in the GC response. Here, we rationally designed T-independent, DNA-origami based virus-like particles (VLPs) with optimal antigenic display of the germline targeting HIV Env immunogen, eOD-GT8, and appropriate T cell help to achieve a potent GC response. In preclinical mouse models, these DNA-VLPs expanded significantly higher frequencies of epitope-specific GC B cells compared with a state-of-the-art clinical protein nanoparticle. Optimized DNA-VLPs primed germinal centers focused on the target antigen and rapidly expanded subdominant broadly neutralizing antibody precursor B cells for HIV with a single immunization. Thus, avoiding scaffold-specific responses augments priming of bnAb precursor B cells, and DNA-VLPs are a promising platform for promoting B cell responses towards challenging subdominant epitopes.

In-depth summary

Claude Sonnet 4.6 · xhigh effort · 7 Nov 2026 · from the full text of version 2 (1 Mar 2025) · CC BY 4.0

DNA origami particles displaying eOD-GT8 with a synthetic T helper epitope gave antigen-focused germinal centers and expanded VRC01-class precursors in humanized mice after one immunization.

Protein scaffolds for multivalent antigen display are themselves thymus-dependent targets, so scaffold-specific B cells may compete with on-target B cells in germinal centers. A DNA origami scaffold has no protein substrate besides the antigen, which allows the paper to test the effect of removing scaffold competitors.

Key results

  1. ShownThe first DNA-VLP design (d40_30mer, 30 antigens) boosted post-boost serum titers but primed germinal centers poorly. Neither DNase I deficiency nor alternative adjuvants rescued this.

    Post-boost endpoint titers were about 1 log above eOD-GT8 monomer. At day 14, GC B and Tfh counts did not exceed monomer, and eOD-specific GC B cells stayed below 1%. DNase I KO mice (n=4/group) matched WT. · d40_30mer DNA-VLPs enhance serum antibody titers…

  2. ShownRaising antigen density to 60 copies on the smaller particle (d30_60mer) increased lectin-pathway complement deposition, FDC capture and antigen-specific GC B cell expansion. The larger 60-copy particle (d40_60mer) did not, so density mattered rather than valency alone.

    MBL binding to d30_60mer was nearly identical to p60mer, while C3 was significantly lower. FDC colocalization at day 4 was seen only for d30_60mer. eOD-specific GC B cells rose about 15-fold in frequency and 10-fold in counts vs monomer. · d40_30mer DNA-VLPs have restricted access to…; Engineering DNA-VLPs with high antigen density…

  3. ShownAdding the PADRE T helper epitope to eOD-GT8 on the 60-copy DNA-VLP raised GC size and made the response far more antigen-specific than with the lumazine synthase protein nanoparticle (p60mer) in C57BL/6 mice.

    eOD-specific GC B cells were 30-40% of GC B cells with d30_60mer-PADRE vs about 5% with p60mer. Their total counts were 2.1-fold higher, although total GC B cells were 4-fold lower than with p60mer (n=8 mice/group). · Synthetic T cell epitopes provide focused T cell…

  4. ShownIn VH1-2 humanized mice, GCs primed by d30_60mer-PADRE were dominated by CD4 binding site-specific B cells, whereas p60mer GCs carried a large lumazine synthase-specific competitor population.

    On-target CD4bs-specific cells were about 60% of GC B cells vs about 20% for p60mer. Lumazine synthase-specific cells were about 13% of the p60mer GC. The on-target to off-target ratio was 25-fold higher with the DNA-VLP (n=8/group). · DNA-VLPs prime epitope-focused germinal centers…

  5. SuggestedSingle-cell BCR sequencing of CD4bs-specific GC B cells at day 14 showed more clonal expansion and enrichment of the 5-amino-acid CDRL3 VRC01-class signature after DNA-VLP priming, but not after p60mer.

    444 (DNA-VLP) and 465 (p60mer) BCR sequences from 2 mice per group. Six public clones carrying the signature were expanded in different DNA-VLP recipients, using mouse IGKV4-61*01. Somatic hypermutation did not differ significantly between groups. · eOD display by DNA-VLPs leads to more efficient…; Figure legends

Methods, data and reproducibility

Icosahedral DNA origami particles were conjugated with eOD-GT8 at varied size, antigen copy number and T helper epitope content. They were compared with eOD-GT8 monomer and a lumazine synthase nanoparticle by lymph node imaging, complement ELISAs, flow cytometry of GC responses and single-cell BCR sequencing in mice.

  • C57BL/6 mice immunized with SMNP adjuvant; serum ELISA, GC flow cytometry, lymph node imaging · mouse · n = n=3 to 8 mice per group; DNase I KO experiments n=4/group
  • VH1-2 humanized mice (human IGHV1-2*02 knock-in) for competitor versus on-target GC B cell probing · mouse · n = n=8 per group (flow)
  • Single-cell BCR sequences of CD4bs-specific GC B cells from VH1-2 mice at day 14 · mouse · n = 444 (d30_60mer-PADRE) and 465 (p60mer) sequences; 2 mice per immunogen
  • Calcium flux in Ramos B cells expressing germline VRC01 BCR · human (Ramos B cell line) · n = 3 independent experiments
Methods & tools
DAEDALUS, ATHENA, Tiamat, SMNP adjuvant, eOD-GT8, Smart-Seq2, MigMAP, PandaSeq, MEGAx, GraphPad Prism v.9, Relion, MotionCor2
Reproducibility gaps
BCR sequences are to be deposited in GenBank with accession numbers still to be determined. Raw imaging data are available only on request. No new code was developed.
Limitations (5)
  • Degradation of DNA-VLPs by extracellular nucleases still needs evaluation for clinical translation. Autoimmune risk from antibody responses to DNA/protein hybrids must be assessed, although no class-switched anti-DNA or anti-VLP antibodies were detected here. · stated by the authors
  • C3 deposition on the optimal DNA-VLP was significantly lower than on the protein nanoparticle, a design element the authors say remains to be improved. · stated by the authors
  • Effects of scaffold responses on re-exposure, such as feedback and imprinting in heterologous boosting, were not tested; only single-dose priming was compared. · stated by the authors
  • The BCR sequencing result rests on 2 mice per immunogen at one timepoint (day 14), so the public-clone and CDRL3 signature findings come from very few animals. · noted in this summary
  • All in vivo work is in mice with one antigen (eOD-GT8) and one protein comparator (lumazine synthase nanoparticle). The attribution of focusing to undiluted T cell help is inferred, not directly tested. · noted in this summary

Read this if you design multivalent nanoparticle vaccines and want data on how scaffold-specific B cells and antigen density shape germinal center priming of rare precursors.

Citations
46
FWCI
133.77
Field percentile
top 1%

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.