Back

Cyclic-thiourea ionizable lipids program multilamellar mRNA-LNPs for sustained expression, frozen-storage stability and potent vaccine immunotherapy

Hui, A.; Cai, R.; Zhang, L.; Huang, Z.; Gu, H.; Zheng, N.; Chen, C.

2026-09-09 bioengineering
10.64898/2026.09.07.749752 bioRxiv
Show abstract

Lipid nanoparticles (LNPs) are the leading delivery vehicles for mRNA therapeutics, yet transient expression and limited formulation stability constrain their broader therapeutic application. Here we report the phenotype-driven discovery of an ionizable lipid (HI-62) bearing a 1,3,5-triazinane-2-thione (cyclic-thiourea) headgroup that reproducibly assemble mRNA into onion-like multilamellar nanoparticles. Cryogenic electron microscopy revealed ordered internal lamellae, while molecular-dynamics simulations supported a cooperative assembly model involving pH-responsive electrostatics, headgroup preorganization and short-range directional interactions. The lead HI-62 formulation maintained greater than 95% mRNA encapsulation across N/P ratios of 3-7, prolonged reporter expression for substantially longer durations than benchmark formulations following intravenous and intramuscular delivery, and retains physicochemical integrity and in vivo potency after repeated freeze-thaw cycles and 18 months of storage at -20. In the tumor models, HI-62 elicits robust cellular responses and achieved significant inhibition of tumor growth. These findings establish cyclic-thiourea headgroup chemistry as a design principle for programming lipid-RNA organization into multilamellar architecture, and provide a design principle for next-generation LNPs platform with durable protein expression and superior stability. The HI-62-based formulation has been utilized in mRNA cancer vaccine program which will enter clinical development in months.

Matching journals

The top 9 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.