A nanoparticle vaccine that targets neoantigen peptides to lymphoid tissues elicits robust antitumor T cell responses

Neoantigens arising from tumor-specific mutations are unique to tumors and attractive vaccine targets, yet peptide vaccines often fail to generate sufficient cytotoxic CD8+ T cell responses. Despite improved in silico prediction of peptide-MHC binding, a key challenge is to deliver antigens to antigen-presenting cells within secondary lymphoid organs to prime robust CD4+ and CD8+ responses. Prior clinical experiences with SLP vaccines show dominant CD4+ responses even when peptides are selected for MHC-I, suggesting suboptimal cross-presentation and CD8+ priming. The study tests the hypothesis that delivering synthetic long neoantigen peptides via cationic lipoplexes to myeloid cells in spleen and lymph nodes will enhance antigen uptake, cross-presentation, and coordinate CD4+ help, thereby boosting CD8+ responses and improving anti-tumor efficacy. The authors focus on recurrent KRAS G12 mutations as model neoantigens and evaluate whether a liposomal SLP platform can overcome prior limitations and synergize with checkpoint blockade.

Vaccination strategies traditionally aim at dendritic cells for priming; however, many neoantigen peptide vaccines predominantly induce CD4+ responses despite MHC-I selection. Effective anti-tumor immunity likely requires coordinated CD4+ help to optimize CTL generation and function within lymphoid organs. Studies suggest vaccines perform best when MHC-I and MHC-II epitopes are colocated within the same region, enabling helper activity. Cytokine cues such as IL-15 can enhance DC activation and co-stimulation for CD8+ priming. Emerging platforms that target lymphoid tissues (e.g., cationic RNA-liposomes, intranodal mRNA, listeria vectors, lymph node-targeted nanoparticles, lipoprotein nanodiscs) have shown improved CD8+ and sometimes CD4+ responses, particularly when combined with checkpoint inhibitors. Nonetheless, peptide vaccines have historically shown modest clinical efficacy, highlighting the need for improved delivery and combination strategies.

Vaccine design and formulation: Synthetic long peptides (SLPs) spanning KRAS G12 mutations (notably G12D1–23) and other neoantigens (e.g., Adpgk, Cop1, CT26-derived 27-mers) were complexed with cationic liposomes to form peptide–lipoplexes (SLP-Lpx). Liposomes were prepared by thin-film hydration using DOTMA, DOPC, and DOPE/DOPS (common ratio 50:30:20) with rhodamine-labeled DOPE for biodistribution where indicated. Films were hydrated with RNase-free water and mixed with peptide at 1:50 (w/w) lipid:peptide, incubated 1 h, and extruded 10× through 200 nm membranes, yielding 100–150 nm particles with measured zeta potential and PDI. CpG ODN 1826 was included as adjuvant in formulations. Cationic formulations were selected over anionic based on superior splenic targeting. Animal models: Female C57BL/6 and BALB/c mice (8–12 weeks) were used under IACUC/AAALAC protocols. Models included: (1) KRAS-G12D genetically engineered mouse (GEM) lung adenocarcinoma induced by intratracheal AdCre; (2) syngeneic subcutaneous MC38 and CT26 tumors engineered to express WT or mutant KRAS (G12D, G12V, G12R). Prophylactic and therapeutic vaccination schedules were tested. T cell depletion (anti-CD4, anti-CD8) evaluated immune dependence. Combinations with anti–PD-1 were assessed therapeutically. Immunization schedules: For peptide alone immunogenicity studies, mice received 100 µg peptide + 50 µg CpG on days 0 and 7; readouts at days 6–8 post-boost. For Lpx studies, ~5 µg peptide dose per injection within 50 µg peptide–Lpx, administered s.c. (also i.v. in screens), with boosts typically weekly (3 total). Neo-Lpx cocktails included KRAS-G12D and additional 27-mer neoantigens. Biodistribution and uptake: Cy5-labeled peptides and rhodamine-labeled liposomes were administered s.c.; organ fluorescence quantified by ex vivo IVIS. Cellular uptake by splenic myeloid populations (CD11b+CD11c− macrophages, CD11c+ DCs, cDCs) measured by flow cytometry (percent uptake, MFI). BMDCs in vitro uptake assays included macropinocytosis inhibition controls. Immunologic assays: IFN-γ ELISpot on splenocytes after re-stimulation with overlapping 9-mer or 15-mer peptides; intracellular cytokine staining, activation/proliferation markers (PD-1, Ki-67), and TIL phenotyping (PD-1, Tim-3, Lag3, CTLA-4). MHC-I dextramers detected antigen-specific CD8+ T cells (e.g., Adpgk-specific). Flow cytometry panels included CD4, CD8, CD11b/c, H2Kb, PD-L1, Foxp3, T-bet, IL-2, IFN-γ, Ki-67. Tumor efficacy: Tumor volumes monitored over time; treatments given prophylactically or therapeutically (post-implantation day 5–14 depending on experiment). In GEM model, weekly G12D-Lpx post-AdCre for 3 weeks. Combination therapy cohorts received anti–PD-1 versus isotype control. Statistical analyses used two-way or repeated-measures ANOVA and t-tests; data reported as mean ± s.e.m.

- Naked KRAS G12D peptide + CpG immunization primarily elicited CD4+ T cell responses; overlapping 15-mers identified CD4+ epitopes (G12D1-9, G12D2-9) with robust IFN-γ ELISpot. - Cationic peptide–lipoplex (G12D1–23-Lpx) improved biodistribution: subcutaneous administration led to a 2–3-fold increase in peptide accumulation in spleen compared with naked peptide. - Myeloid uptake enhancement: percentages of splenic macrophages and DCs taking up peptide were markedly higher with Lpx; mean fluorescence intensity of Cy5-peptide uptake in macrophages and DCs increased up to ~100-fold versus naked peptide. - Immunogenicity shift: Peptide-Lpx vaccination elevated antigen-specific CD8+ T cell responses while maintaining strong CD4+ responses, converting responses from predominantly CD4+ to balanced CD4+/CD8+. - Tumor control in syngeneic models: Prophylactic G12D-Lpx significantly inhibited MC38-G12D tumor growth; WT KRAS Lpx had no effect. Therapeutic vaccination reduced growth rate but tumors continued to progress, indicating partial control. - GEM lung model: Weekly G12D-Lpx post-transformation reduced tumor volume and burden, increased hematopoietic/T cell infiltrates, and reduced PD-1 expression in tumors. - CD8+ dependence: Tumor growth inhibition by G12D-Lpx was abrogated by CD8+ T cell depletion; CD4+ depletion had minimal impact, indicating CD8+ T cells mediate efficacy. - Generalizability to other neoantigens: Adpgk and Cop1 27-mer neoantigens delivered as Lpx elicited both CD4+ and CD8+ responses and enhanced control in CT26/MC38 contexts; multiple CT26 neoantigens in Neo-Lpx suppressed tumor growth more profoundly than peptide alone. - Checkpoint combination: Neo-Lpx combined with anti–PD-1 produced markedly greater tumor suppression than either monotherapy, yielding partial complete responses in 6/10 mice. - TIL phenotype: Vaccination increased antigen-specific CD8+ TILs but also upregulated exhaustion markers (PD-1, Tim-3, Lag3, CTLA-4) on TILs; tumors upregulated PD-L1 and MHC-I after vaccination.

The study addresses the limitation of SLP-based cancer vaccines that predominantly induce CD4+ T cells and show limited clinical benefit by improving antigen delivery to secondary lymphoid tissues. Cationic lipoplex delivery enhances uptake by splenic myeloid cells and cross-presentation, thereby driving robust CD8+ cytotoxic responses while preserving CD4+ helper responses. This dual activation translates into significant tumor growth inhibition in KRAS-G12D-driven syngeneic and GEM models. However, despite increased CD8+ TILs, monotherapy often leads to T cell dysfunction with upregulation of inhibitory receptors and PD-L1 in tumors, limiting efficacy. Combining the lipoplex neoantigen vaccine with PD-1 blockade overcomes some exhaustion-related barriers, substantially improving tumor control and inducing partial complete responses in a subset of animals. These findings reinforce the concept that effective neoantigen vaccination requires both targeted delivery to lymphoid APCs to generate balanced CD4+/CD8+ responses and combinatorial checkpoint inhibition to sustain T cell function within the tumor microenvironment.

Encapsulating synthetic long neoantigen peptides in cationic liposomes targets antigens to splenic myeloid cells, enhancing processing and presentation to elicit coordinated CD4+ and CD8+ T cell responses. This approach improves antitumor efficacy against KRAS-G12D in both syngeneic and GEM mouse models and synergizes with PD-1 checkpoint blockade to produce profound tumor suppression, including partial complete responses. The platform preserves the safety and manufacturability advantages of peptide vaccines while overcoming a key limitation—weak CD8+ priming. Future work should optimize formulations and dosing, expand neoantigen repertoires to match diverse MHC class I/II alleles, and pair with rational combinations (multi-checkpoint blockade or agents reversing T cell exhaustion) to maximize durability and breadth of responses, with a view toward clinical translation.

- Monotherapy vaccination induced TIL exhaustion (PD-1+, Tim-3+, Lag3+, CTLA-4+), limiting complete tumor rejection; vaccine efficacy depended on combination with PD-1 blockade in established tumors. - Evidence is preclinical in mouse models (C57BL/6, BALB/c; MC38, CT26; KRAS-G12D GEM), and translatability to humans and diverse HLA alleles remains to be established. - Single/limited antigen targeting risks tumor immune escape; although multi-neoantigen cocktails were tested, breadth may still be insufficient for heterogeneous tumors. - While cationic liposomes improved splenic targeting, optimal biodistribution parameters and long-term safety profiles were not fully characterized. - Therapeutic settings showed slowed but continued tumor growth with vaccine alone, highlighting incomplete efficacy without combination therapy.