Technical Review: A modular mRNA–LNP vaccine platform enables integrated RNA, lipid and antigen design to protect against CCHFV

Authors: Touraj Farzani, Nallely Espinoza, Mahdiyeh M. Manafi, Stephen R. Welch, JoAnn D. Coleman-McCray, Virginia Aida-Ficken, Jessica R. Spengler, Éric Bergeron, Christina F. Spiropoulou, Clarice Borges, Kristine Bielecki, Lee Li, Devan Shah, Marietou Paye, Eugenia Rojas, Sophie N. Spector, Pedram Samani, Lisa E. Hensley, Al Ozonoff, and Pardis C. Sabeti

Affiliations: Massachusetts Institute of Technology, Harvard University, and University of Massachusetts Boston; Boston, MA, USA

Background Introduction

Crimean–Congo hemorrhagic fever virus (CCHFV) is a high-consequence emerging tick-borne pathogen with severe disease potential and no licensed vaccine. A successful vaccine must yield predictable immune polarization and tolerability, with "plug-and-play" design a highly desired bonus. mRNA-lipid nanoparticle (LNP) vaccines are a natural solution due to their high programmability, itself a consequence of their multi-component design.

This paper quantitatively maps how lipid identity, UTR configuration, and 5′ cap structure shape early innate cytokine landscapes during antigen presentation, and how those landscapes translate into downstream T cell phenotypes and antibody quality. mRNA-LNP vaccine design is therefore evaluated as an emergent system, where a given antigen can yield different immune trajectories based on changes to either the LNPs or their mRNA payload.

Fig 1. Schematic overview of the integrated mRNA–LNP engineering framework spanning antigen architecture, ionizable lipid chemistry, and mRNA regulatory design. Two CCHFV immunogen classes were evaluated: the intracellular nucleocapsid (NP) and the secreted glycoprotein complex (sGCs; mucin-like domain (MLD)–GP38). Targeted antigen modifications included removal of a canonical caspase-3 cleavage motif in NP (NPmut) and deletion of two conserved GP38 N-linked glycosylation sites (GP38Δglyc; N129 and N179). LNP formulation incorporated the ionizable lipid BP-104. mRNA design variables included 5′ cap structure (Cap0 ARCA; Cap1 AG; Cap1 CleanCap M6), untranslated regions (UTRs; viral/CCHFV, T, and ACTA-1), and modified nucleotides (N1-methylpseudouridine-UTP and 5-methyl-CTP). Ionizable lipid chemistry sets the innate cytokine/chemokine context, UTR architecture tunes T cell cytokine polarization and cellular distribution, cap structure scales response magnitude and antibody output, and antigen architecture defines the protective ceiling.

Materials and Methodology

The researchers used a platform combining a modular antigen panel with systematic variation in mRNA regulatory architecture. LNPs with various ionizable lipids were prepared with microfluidic mixing and evaluated for antigen presentation in vitro. Finally, an in vivo test was conducted on a murine model, where single-antigen and combination CleanCap M6 regiments were compared over 14 days.

Results

Lipid and physicochemical benchmarking

BP‑104 LNPs consistently form monodisperse particles (PDI <0.15) with near-neutral zeta potential (~−8 mV), ~90% encapsulation, and apparent pKa values of 6.6–6.9. Furthermore, they show minimal cytotoxicity and liver histopathology, comparable to SM‑102 and ALC‑0315. This indicates that BP‑104 is a well‑behaved ionizable lipid suitable for mechanistic studies, decoupling delivery efficiency from downstream differences in immune programming.

Innate pathway activation

In reporter assays, NPmut LNPs containing BP‑104, SM‑102, or ALC‑0315 all elicit robust NF‑κB, IL‑6, and type I IFN signaling. ALC‑0315 tends to give the highest IFN‑I at some doses, with BP‑104 intermediate. This data demonstrates that ionizable lipids themselves are active innate stimuli and that BP‑104 can match clinically used lipids in canonical pathway engagement without excessive overactivation.

UTR and lipid control of DC–T priming

DC–T co-cultures reveal that UTR architecture is a practical tuning knob whose effect depends on antigen class: NP mRNAs show relatively stable CD8⁺ IFN‑γ/TNF‑α across most UTRs but variable CD4⁺ responses (including IL‑4/IL‑17A in some contexts), whereas sGCs mRNAs show marked UTR sensitivity in CD8⁺ IFN‑γ/IL‑2 and mixed Th1/Th2/Th17 CD4⁺ profiles. Lipid identity imprints distinct DC cytokine/chemokine milieus and T‑cell fates, with ALC‑0315 driving the most inflammatory, Th1/Th17‑biased responses, SM‑102 the weakest, and BP‑104 an intermediate, more balanced profile.

Fig 2. mRNA construct configurations used for in vivo studies, grouped by cap-structure experiments. ARCA-capped constructs were paired with alternative UTRs (viral/CCHFV, T, ACTA-1), whereas AG- and CleanCap M6–capped constructs used defined 5′/3′ UTR configurations (CCHFV-N and CCHFV-M) and a segmented poly(A) tail. Immunogenicity studies used intramuscular vaccination with 10 μg mRNA/LNP per mouse for single-antigen groups and 20 μg total mRNA/LNP for combination groups. Protective efficacy was evaluated in transiently immunosuppressed C57BL/6 mice in a lethal CCHFV challenge model.

In vivo cellular polarization and regulatory architecture

In ARCA‑capped NP vaccines, UTR choice strongly reshapes NP‑specific cellular immunity. CCHFV and ACTA‑1 UTRs enhance IL‑2 ELISPOT responses relative to T‑UTR, and ACTA‑1 broadens CD4⁺ outputs across IFN‑γ, IL‑2, IL‑4, TNF‑α, and IL‑17A, while CD8⁺ profiles remain comparatively similar. Under AG capping, NP/NPmut induce Th1‑biased IFN‑γ responses whereas sGCs constructs are IL‑4‑dominant; CleanCap M6 predominantly scales response magnitude (boosting NP/NPmut CD8⁺ IFN‑γ and sGCs(GP38Δglyc) CD4⁺ IL‑2/TNF‑α) without altering antigen-imposed polarization, positioning the cap as a quantitative gain control.

Tfh/GC organization and humoral responses

Draining lymph node analyses show that NPmut preferentially expands Tfh cells, whereas sGCs(GP38Δglyc) induces the strongest antigen-specific GC B‑cell responses, indicating that antigen structure independently governs Tfh vs GC recruitment. Serology and neutralization assays reveal that ARCA constructs elicit weak, strain-dependent antibodies, AG capping preserves antigen-specific hierarchies (NP seronegative, NPmut low, sGCs strong), and CleanCap M6 increases titers across all antigens while maintaining IgG1‑dominant sGC responses; sGCs(GP38Δglyc) yields higher neutralization titers than sGCs, consistent with improved exposure of neutralizing epitopes.

Protective efficacy in lethal challenge

In the stringent lethal CCHFV model, CleanCap M6 sGCs monotherapy achieves 90% survival with minimal clinical signs, sGCs(GP38Δglyc) achieves ~70% survival, NP and NPmut confer no protection, and NP+sGCs combinations give intermediate, non-additive protection that never exceeds sGCs alone. Protective efficacy tracks best with sGCs-driven GC B‑cell responses and neutralizing antibody quality rather than with total T‑cell magnitude, underscoring that glycoprotein-targeted humoral immunity is the key effector axis for rapid CCHFV control.

Conclusion

Farzani and colleagues showcased BP-104 as a benchmark-comparable ionizable lipid. When integrated in optimized CleanCap M6 sGCs mRNA-LNPs, it could achieve robust protection against CCHFV. This reinforces the emergent properties of mRNA-LNP design, which represent a key aspect of their continued relevance in varied vaccine applications.

This paper utilized the PreciGenome NanoGenerator Flex-M for LNP synthesis. Uniform synthesis conditions are essential for screening RNA-LNP design without complicating factors from preparation, and the Flex-M's microfluidic mixing made excellent LNP uniformity possible in these experiments.

For further details, check out the link to the article here:

https://www.biorxiv.org/content/10.64898/2026.01.17.699915v1.article-info