Degradable branched (DB-)lipidoids built by one-pot tandem epoxide ring-opening and acyl-chloride esterification are an easily synthesized class of ionizable lipids for lipid-nanoparticle (LNP) mRNA delivery, for which three empirical structural criteria — total tail carbon number of 18 (the "18-Carbon Rule"), tail symmetry of 1, and a diamine headgroup combining one primary amine with one dimethylamino, diethylamino, or pyrrolidinyl tertiary amine — were shown in 2024 to predict hepatic mRNA delivery potency in mice. Three issues remain open: whether these criteria generalize when the tail chemistry is extended from saturated, linear epoxides and acyl chlorides to unsaturated or branched ones; how potency changes when headgroups beyond two-body-tail amines (polyamines, dendrimers bearing multiple body tails) are used; and why DB-LNPs that are potent in vivo perform poorly in the HepG2 (human hepatocellular carcinoma cell) in vitro screen (and vice versa). Solving these would deliver generalizable, prospectively testable design rules for degradable branched lipidoids and a mechanistically grounded screening strategy for predicting in vivo mRNA delivery.

Background

Messenger RNA (mRNA) therapeutics and gene editors require delivery vehicles that protect the RNA, target the right tissue, and release it into the cytoplasm; the clinically dominant platform is the lipid nanoparticle (LNP), a formulation whose functional core is an ionizable lipid — an amine-bearing lipid that is neutral in blood but protonates in the acidic endosome to aid escape. Lipidoids are lipid-like ionizable materials discovered through combinatorial chemistry, typically by reacting amines with epoxides or acyl chlorides, which allows large libraries to be synthesized and screened. A persistent goal of the field is to replace purely empirical screening with structural design rules that predict in vivo potency.

Two structural motifs motivate this problem. First, branched hydrophobic tails had already been shown to enhance LNP performance: a lipidoid differing from its analogue only by a one-carbon branch in its tail was about ten times more potent in vivo, attributed to stronger surface ionization at late-endosomal pH (Hajj et al., Small 15, 1805097, 2019), and branching was further linked to liposomal stability, fusogenicity, and mRNA delivery (Hashiba et al., Small Science 3, 2200071, 2023). Second, most clinically used ionizable lipids are biostable; degradable variants with ester linkages clear more easily but were historically harder to synthesize in branched form. Han et al. (Nature Communications 15, 1762, 2024) introduced a tandem, one-pot construction: an amine headgroup first ring-opens an epoxide (giving an aminoalcohol bearing a "body" tail), then acyl chlorides attach ester-linked "branch" tails, yielding degradable branched lipidoids (DB-lipidoids) denoted x-m-n (amine x; body tail of m carbons; branch tail of n carbons). Screening Library 1 (25 tail combinations) and Library 2 (20 amines) both in HepG2 cells (a human hepatocellular carcinoma cell line widely used for in vitro LNP screening) and in C57BL/6 mice (a standard inbred laboratory strain; luciferase-mRNA LNPs, 0.1 mg/kg, bioluminescent imaging) produced three structural criteria for potent DB-lipidoids: (1) total tail carbon number m+n=18m+n=18 (the "18-Carbon Rule", necessary but not sufficient); (2) tail symmetry n/(m−2)=1n/(m-2)=1; (3) a diamine headgroup with one primary amine and one dimethylamino, diethylamino, or pyrrolidinyl tertiary amine separated by two or three methylene units. The criteria correctly predicted the (weaker) performance of four held-out analogs, and the lead 11-10-8 outperformed the benchmark ionizable lipid DLin-MC3-DMA (MC3, the ionizable lipid of the clinically approved siRNA drug patisiran) about six-fold in hepatic mRNA delivery, enabling hepatic gene editing of TTR (transthyretin, the protein whose hepatic gene is targeted in hereditary transthyretin amyloidosis) and protein-supplementation therapy with FGF21 (fibroblast growth factor 21, a protein hormone) mRNA in mice.

Three limitations were explicitly acknowledged and remain unresolved. First, all Library-1 tails came from saturated, linear epoxides and acyl chlorides, so the criteria are unknown outside that chemistry, even though the construction method tolerates unsaturated or branched monomers. Second, all Library-2 headgroups attach exactly two body tails, so DB-lipidoids with polyamine or dendrimer headgroups bearing multiple body tails have never been evaluated. Third, in both libraries the in vitro (HepG2) and in vivo (mouse) potency rankings were essentially uncorrelated — potent in vivo DB-LNPs were typically inferior to MC3 in vitro — and the reason was stated to be unclear. This in vitro/in vivo mismatch is not unique to DB-lipidoids: a direct comparison of hundreds of nanoparticles found only a weak correlation between in vitro and in vivo delivery (Paunovska et al., Nano Letters 18, 2148–2157, 2018), making reliable in vitro prescreening a field-wide unsolved problem.

Problem Statement

Starting from the degradable branched lipidoid (DB-lipidoid) platform of Han et al. (Nature Communications 15, 1762, 2024) — ionizable lipids built by tandem amine–epoxide ring-opening followed by acyl-chloride esterification, in which an amine headgroup carries two body tails each bearing an ester-linked branch tail — determine whether its empirically derived structural criteria for in vivo mRNA delivery potency generalize beyond the chemistry they were derived from, and explain why in vitro screening fails to predict in vivo potency for this class. Concretely, working from the published DB-lipidoid structure–activity data and the cited studies of unsaturated and branched ionizable-lipid tails: (1) predict whether total tail carbon number 18 and tail symmetry n/(m−2)=1n/(m-2)=1 remain predictive when the tail monomers are unsaturated or branched epoxides and acyl chlorides rather than only saturated, linear ones, stating the structure–activity analysis and the minimal set of held-out analogs with their in vivo (luciferase-mRNA LNP, mouse) measurements that would confirm or refute the prediction; (2) predict how multi-body-tail architecture — headgroups such as polyamines or dendrimers that attach more than two body tails — affects in vivo mRNA delivery potency relative to the two-tail criterion, on the basis of the available headgroup and lipidoid data, stating the lipidoid series and the in vivo (luciferase-mRNA LNP, mouse) comparison against the corresponding two-body-tail lipidoids that would confirm or refute the prediction; and (3) identify the biological and/or physicochemical mechanism(s) responsible for the discrepancy between in vitro (HepG2, a human hepatocellular carcinoma cell line) and in vivo (mouse) potency rankings of DB-LNPs — several LNPs weak in vitro are potent in vivo and vice versa — establishing either an in vitro readout (or corrected in vitro protocol) whose ranking matches in vivo potency on the available data or a demonstrated reason that no available readout can, together with the readout or protocol change that would need to be tested. All three components are required. The common objective of all three components is a set of experimentally testable, generalizable design criteria for degradable branched lipidoids that predict in vivo mRNA delivery potency, usable without relying on the currently unreliable in vitro prescreen.

References

  1. Han, X.; Xu, J.; Xu, Y.; Alameh, M.-G.; Xue, L.; Gong, N.; El-Mayta, R.; Palanki, R.; Warzecha, C. C.; Zhao, G.; Vaughan, A. E.; Wilson, J. M.; Weissman, D.; Mitchell, M. J. In situ combinatorial synthesis of degradable branched lipidoids for systemic delivery of mRNA therapeutics and gene editors. Nature Communications 15, 1762 (2024). DOI: 10.1038/s41467-024-45537-z. https://doi.org/10.1038/s41467-024-45537-z
  2. Hashiba, K.; Sato, Y.; Taguchi, M.; Sakamoto, S.; Otsu, A.; Maeda, Y.; Shishido, T.; Murakawa, M.; Okazaki, A.; Harashima, H. Branching ionizable lipids can enhance the stability, fusogenicity, and functional delivery of mRNA. Small Science 3, 2200071 (2023). DOI: 10.1002/smsc.202200071. https://doi.org/10.1002/smsc.202200071
  3. Hajj, K. A.; Ball, R. L.; Deluty, S. B.; Singh, S. R.; Strelkova, D.; Knapp, C. M.; Whitehead, K. A. Branched-tail lipid nanoparticles potently deliver mRNA in vivo due to enhanced ionization at endosomal pH. Small 15, 1805097 (2019). DOI: 10.1002/smll.201805097. https://doi.org/10.1002/smll.201805097
  4. Paunovska, K.; Sago, C. D.; Monaco, C. M.; Hudson, W. H.; Gamboa Castro, M.; Rudoltz, T. G.; Kalathoor, S.; Vanover, D. A.; Santangelo, P. J.; Ahmed, R.; Bryksin, A. V.; Dahlman, J. E. A direct comparison of in vitro and in vivo nucleic acid delivery mediated by hundreds of nanoparticles reveals a weak correlation. Nano Letters 18, 2148–2157 (2018). DOI: 10.1021/acs.nanolett.8b00432. https://doi.org/10.1021/acs.nanolett.8b00432
  5. Han, X.; Xu, Y.; Ricciardi, A. S.; Xu, J.; Xiang, Y.; Palanki, R.; Chowdhary, V.; Xue, L.; Gong, N.; Alameh, M.-G.; Peranteau, W. H.; Wilson, J. M.; Reker, D.; Weissman, D.; Mitchell, M. J. Plug-and-play assembly of biodegradable ionizable lipids for potent mRNA delivery and gene editing in vivo. Proceedings of the National Academy of Sciences 123, e2528144123 (2026). DOI: 10.1073/pnas.2528144123. https://doi.org/10.1073/pnas.2528144123
  6. Padilla, M. S.; Mrksich, K.; Wang, Y.; Haley, R. M.; Li, J. J.; Han, E. L.; El-Mayta, R.; Kim, E. H.; Dias, S.; Gong, N.; Teerdhala, S. V.; Han, X.; Chowdhary, V.; Xue, L.; Siddiqui, Z.; Yamagata, H. M.; Kim, D.; Yoon, I.-C.; Wilson, J. M.; Radhakrishnan, R.; Mitchell, M. J. Branched endosomal disruptor (BEND) lipids mediate delivery of mRNA and CRISPR-Cas9 ribonucleoprotein complex for hepatic gene editing and T cell engineering. Nature Communications 16, 996 (2025). DOI: 10.1038/s41467-024-55137-6. https://doi.org/10.1038/s41467-024-55137-6
  7. Xue, Z.; Lummerstorfer, M.; Brunner, L.; Lächelt, U. Partially epoxylated lipidoids with tertiary and secondary amines for messenger RNA transfection. ChemistryEurope 4, e202500429 (2026). DOI: 10.1002/ceur.202500429. https://doi.org/10.1002/ceur.202500429
  8. Lummerstorfer, M.; Xue, Z.; Zheng, D.; König, L.; Brunner, L.; Lächelt, U. Lipid nanoparticles for Cas9 ribonucleoprotein delivery: design and evaluation of ionisable oligoamine-lipidoids. European Journal of Pharmaceutical Sciences 224, 107596 (2026). DOI: 10.1016/j.ejps.2026.107596. https://doi.org/10.1016/j.ejps.2026.107596
  9. Ge, X.; He, Z.; Yang, H.; Pan, X.; Yan, L.; Shi, Y.; Chen, Y.; Liu, Z. Impact of tail unsaturation in ionizable lipids on mRNA delivery efficiency and immunogenicity of lipid nanoparticles. Journal of Controlled Release 384, 113906 (2025). DOI: 10.1016/j.jconrel.2025.113906. https://doi.org/10.1016/j.jconrel.2025.113906
  10. Samaridou, E.; Simon, J.; Beck-Broichsitter, M.; Davidson, G.; Levkin, P. A. Rational design of unsaturated, thioether ionizable lipids for enhanced in vivo mRNA delivery. Advanced Healthcare Materials 14, 2501037 (2025). DOI: 10.1002/adhm.202501037. https://doi.org/10.1002/adhm.202501037
  11. Strelkova Petersen, D. M.; Weiss, R. M.; Hajj, K. A.; Yerneni, S. S.; Chaudhary, N.; Newby, A. N.; Arral, M. L.; Whitehead, K. A. Branched-tail lipid nanoparticles for intravenous mRNA delivery to lung immune, endothelial, and alveolar cells in mice. Advanced Healthcare Materials 13, 2400225 (2024). DOI: 10.1002/adhm.202400225. https://doi.org/10.1002/adhm.202400225
  12. Jozić, A.; Le Roux, C.; Kim, J.; Berchel, M.; Sahel, D. K.; Bodi, E. K.; Palumbo, M.; Vasudevan, A.; Murthy, N. T. V.; Eygeris, Y.; Gautam, M.; Bloom, E.; Barnes, A. P.; Jaffrès, P.-A.; Sahay, G. In vivo endosomal escape assay identifies mechanisms for efficient hepatic LNP delivery. Nature Biotechnology (2026). DOI: 10.1038/s41587-026-03022-6. https://doi.org/10.1038/s41587-026-03022-6
  13. Xu, H.; Fan, J.; Liu, J.; Wei, Q.; Li, R.; Yuan, P.; Xiao, B.; Piao, Y.; Sun, W.; Xiang, J.; Shao, S.; Zhou, Z.; Shen, Y.; Slater, N. K. H.; Tang, J. Tetra-branched ionizable lipids enhance the stability and transfection efficiency of lipid nanoparticles for mRNA delivery. Chinese Chemical Letters 36, 110833 (2025). DOI: 10.1016/j.cclet.2025.110833. https://doi.org/10.1016/j.cclet.2025.110833
  14. Kimura, S.; Tsutsumi, S.; Nakamura, N.; Tsujihata, S.; Lyu, F.; Nakashima, Y.; Kita, S.; Maenaka, K.; Hashiya, F.; Inagaki, M.; Abe, N.; Kimura, Y.; Abe, H. A branched ionizable lipid nanoparticle platform for versatile in vivo delivery of nucleic acids: Validation from mRNA to capped circular RNA. Cell Biomaterials, 100555 (2026). DOI: 10.1016/j.celbio.2026.100555. https://doi.org/10.1016/j.celbio.2026.100555