Cargo-Agnostic Lipids, Cargo-Specific Safety: What LNP Optimization Actually Shows for Embryonic CRISPR Delivery

Question: How have lipid nanoparticle formulations been optimized for delivering CRISPR/Cas9 ribonucleoproteins versus mRNA cargos to embryonic tissues in vivo, and what biodistribution and toxicity tradeoffs distinguish these cargo types?

Answer

The retrieved literature does not contain a head-to-head optimization of LNP-RNP versus LNP-mRNA in embryonic tissue; it contains one in utero LNP genome-editing study [14] and a broader in vivo literature that treats cargo format as a secondary variable behind lipid composition. Taken together, these papers support a specific position: LNP optimization has converged on lipid-level solutions — ionizable lipid pKa and tail unsaturation, PEG-lipid shedding, N/P ratio, and organ-targeting helper lipid ratios — that are largely cargo-agnostic once a size envelope is respected [1] [6] [11], whereas the RNP-versus-mRNA distinction expresses itself less in biodistribution than in (i) internal particle structure and loading homogeneity constraints unique to nucleic acid cargo [5] [7], and (ii) innate-immune, genomic-stability and repeat-dosing liabilities that favor RNP in high-stress delivery contexts [3] [2]. Both formats remain liver-tropic by default, and no retrieved paper shows either format escaping hepatic default in the fetus.

In utero LNP editing exists, but only where fetal anatomy supplies the targeting

The single demonstration of LNP-delivered CRISPR editing in utero exploits developmental hematopoiesis: fetal hematopoietic stem cells reside in the fetal liver, a tissue LNPs access efficiently, and in utero delivery of LNPs carrying Cre recombinase or CRISPR-Cas9 components edits bona fide long-term repopulating HSCs while preserving multilineage reconstitution [14]. Efficiency is stage-dependent rather than formulation-limited, with greater HSC editing during fetal liver hematopoiesis than after birth, and an LNP formulation was identified that improves neonatal HSC targeting without antibody-mediated targeting [14]. Cargo-format boundaries are already visible in that system: Cas9 was delivered by LNP while the repair template was delivered by AAV to achieve in vivo homology-directed repair in multiple tissues, i.e. the LNP carried the nuclease and not the full editing payload [14].

The developmental window is currently defined by viral, not lipid, vectors

The midgestational access problem — reaching fetal circulation and CNS while limiting peripheral organ exposure — has been solved by surgical uterine-window access to the vitelline vein at E12.5 with AAV, not LNP [15]. In that platform, tropism is developmental-stage-dependent, with AAV9 showing higher CNS penetrance and lower liver transduction at E12.5 than at E15.5 [15]. This matters for the question as posed: the gestational timing rules that any LNP-RNP versus LNP-mRNA comparison would need to respect have been established for a different vector class, and have not been transferred to LNPs.

Formulation optimization has been pushed toward cargo-size tolerance rather than cargo-format specialization

A single commercially available LNP formulation delivered cargos spanning more than 300-fold in size, from ~20 nt antisense oligonucleotides to ~6.7 kb CRISPR editors including Cas9 nucleases, adenine base editors and CRISPRa systems, with functional dystrophin restoration and utrophin upregulation [1]. Protein cargo is accommodated by the same composition-tuning logic: enveloped-virus-inspired ionizable lipid designs yielded LNPs that functionally deliver CRISPR-Cas ribonucleoprotein complexes to mammalian cells [13], and a condensate platform was engineered to recruit both Cas9 RNP and mRNA through the same vehicle [18]. Where the problem is instead potency, optimization targets the ionizable lipid itself — engineering unsaturated thioether lipids with more hydrophobic, less protonatable headgroups and longer tertiary-amine alkyl groups improved in vivo mRNA delivery more than 200-fold while matching the safety of an approved benchmark [6].

mRNA cargos carry internal-structure constraints that RNP cargos do not

mRNA loading level, not only lipid composition, determines functional output: LNPs with the highest mRNA loading levels were the least transfection competent after ultracentrifugation fractionation [5]. Internal organization and cargo homogeneity are separately actionable, with N/P=6 giving more uniform mRNA copy number per particle than N/P=3 and higher transfection efficacy [11], and a more ordered internal structure correlating with in vitro activity while lyophilization induces mRNA–lipid phase separation and activity loss [7]. For mRNA specifically, the chemical modification requirement is route-dependent: unmodified mRNA performed comparably to N1-methylpseudouridine-modified mRNA when delivered by a non-endosomal route, but modification was necessary for high expression and low inflammatory response via an endosomal route [17].

The RNP-versus-mRNA tradeoff is immunogenicity and genome stability, not biodistribution

Direct comparison across mRNA and RNP formats within one study shows the tradeoff explicitly: lipid-based delivery achieved the highest editing rates but at the cost of genomic instability risks, interferon pathway activation and a pro-inflammatory shift in paracrine activity, while RNP delivery reduced editing rates in high-stress systems and improved them in lower-stress modalities such as cell-penetrating peptides and electroporation [3]. That stress-dependence is why cargo-agnosticism cannot be assumed: format and delivery method interact, and editing efficiency alone is an insufficient selection metric [3]. In vivo, repeat dosing of Cas9 LNPs elicited a strong adaptive immune response yet still produced ≥35% editing in skeletal muscle, indicating that immunogenicity and potency can be partly decoupled by formulation and tissue [2]. Lipid chemistry can also relieve the toxicity side of the trade directly: ionizable lipids bearing a ROS-scavenging sulfide motif showed a 3.3-fold shorter hepatic half-life and 29-fold lower total hepatic exposure than MC3 with negligible systemic, hepatic or hematological toxicity on repeated non-human primate dosing, while still enabling 57% editing in retinal pigment epithelium [20].

Biodistribution is set by lipid composition and protein corona, and defaults to liver for both formats

LNP tropism is a composition property: high-throughput in vivo screening of 122 barcoded-mRNA LNPs identified hepatic and extrahepatic formulations and linked transfection to single-particle protein corona characterization [16], and a 25-formulation C14120 library with DNA-barcoded pooled readout found organ-specific targeting, with lung specificity validating but spleen specificity not reproducing on individual retest [19]. Composition rules can decouple on-target from off-target exposure, as an intraperitoneal 48-formulation map raised tumor luciferase signal more than 6-fold over library median while cutting liver exposure ~60%, with no treatment-related liver or lung toxicity on histopathology [12]. Acid-degradable PEG-lipid LNPs achieved superior cardiac transfection with reduced off-target liver uptake relative to a standard formulation [9]. The baseline to beat is hepatic: LNPs default to liver tropism [8], and LNPs accumulated in liver with measurable transcriptomic perturbation where cardiac-progenitor extracellular vesicles did not [10]. Placental delivery is a distinct barrier case, where LNP-mediated mRNA delivery to syncytiotrophoblasts increased translation and secretion even under hypoxia [4].

Conflicts and gaps

The central conflict is that cargo-agnosticism and format-specificity are both supported. One formulation delivers 20 nt to 6.7 kb cargos [1] while format and delivery method interact in a stress-dependent manner [3]; these are not reconciled by any retrieved paper. The embryonic evidence base is a single study in fetal liver HSCs [14], and it does not report fetal biodistribution, fetal toxicity, germline editing, or offspring outcomes — none of which any retrieved LNP paper reports for embryonic tissue. No retrieved study compares LNP-RNP and LNP-mRNA at matched editing dose in any embryonic or fetal tissue, and no retrieved study performs barcoded LNP library screening with fetal organ readout, though the methodology exists for adult animals [16] [19]. Whether the mRNA-specific liabilities (loading heterogeneity, ordered-core dependence, interferon activation) are more or less consequential in rapidly dividing embryonic cells than in adult tissue is entirely untested [5] [11] [3].

What would change the answer

A paired-dose experiment injecting an LNP-RNP and an LNP-mRNA formulation matched on editing outcome at E12.5–E15.5, with fetal organ-level biodistribution by DNA barcoding and offspring follow-through, would directly resolve the format question the retrieved literature leaves open [14] [16]. Applying the multiomic readout already used to compare mRNA and RNP formats across delivery modalities [3] to fetal liver and CNS after in utero dosing would establish whether the lipid-associated interferon and genomic-instability costs persist in embryonic cells. Screening barcoded LNP libraries with fetal-tissue readout, as done in adults [16] [19], would test whether fetal liver tropism is the ceiling or a formulation artifact. Finally, testing rapidly clearing, ROS-scavenging ionizable lipids [20] in pregnancy would determine whether the potency–toxicity decoupling demonstrated in non-human primates holds when the exposure window includes a developing organism.

References

[1] Galbiati, P., Leclerc, D., Mombled, M., Khan, R., Ralu, M., Bimbi, G., Scalisi, G., Mamchaoui, K., Tedesco, F. S., Albini, S., Amendola, M.. "A Single Lipid Nanoparticle Formulation Enables Delivery of Diverse RNA Therapeutics to Human Muscle Models of Duchenne Muscular Dystrophy". bioRxiv. 2026-07-29T00:00:00.000Z. URL: https://doi.org/10.64898/2026.07.29.741396

[2] Iyer, S., Daman, K., Sun, Y., Tutto, A., Holbrook, S., Joynt, A., Yan, J., Ambegaokar, P., Guo, D., Liu, P., Stauffer, J., Lee, S., Gallagher, T., Cox, G. A., Keeler, A. M., Siegwart, D. J., Emerson, C. P., Wolfe, S. A.. "Efficient Cas9 nuclease-based editing in skeletal muscle via lipid nanoparticle delivery". bioRxiv. 2025-05-12T00:00:00.000Z. URL: https://doi.org/10.1101/2025.05.12.653518

[3] Graham, J. P., Arteaga, A. V., Moghaddam, A. S., Spiller, K. L., Laverty, D. J., Gonzalez-Fernandez, T.. "Integrative Multiomic Analysis Reveals How Non-Viral Delivery System Selection Shapes CRISPR Gene Editing Outcomes in Stem Cells.". bioRxiv. 2026-05-30T00:00:00.000Z. URL: https://doi.org/10.64898/2026.05.29.728905

[4] Young, R. E., Vijayakumar, T., Darji, K., Hofbauer, S. I., Alameh, M.-G., Weissman, D., Riley, R. S.. "Investigating the Impact of Hypoxia and Syncytialization on Lipid Nanoparticle-Mediated mRNA Delivery to Placental Cells". bioRxiv. 2025-06-25T00:00:00.000Z. URL: https://doi.org/10.1101/2025.06.23.661191

[5] Suiyang Liao, Shuangyu Wang, Abishek Wadhwa, Alex Birkenshaw, Kevin Fox, Miffy Hok Yan Cheng, Irafasha C. Credo, Armando Alcazar Magana, Nuthan Vikas Bathula, Chiao Hao Ho, Melody Cheng, Leonard J. Foster, Kenneth W. Harder, Colin J.D. Ross, Pieter R. Cullis, Anna K. Blakney. "The transfection potency of lipid nanoparticles containing mRNA depends on relative loading levels". chemrxiv. 2024-11-14T05:43:00.000Z. URL: https://doi.org/10.26434/chemrxiv-2024-l9fn9

[6] 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". bioRxiv. 2024-12-30T00:00:00.000Z. URL: https://doi.org/10.1101/2024.12.30.630759

[7] Chen, X., Fang, J., Ge, X., Li, M., Jiang, F., Hong, L., Liu, Z.. "Revealing a Correlation between Structure and in vitro Activity of mRNA Lipid Nanoparticles". bioRxiv. 2026-03-26T00:00:00.000Z. URL: https://doi.org/10.1101/2024.09.01.610730

[8] Marius Kausas, Valentina Schastlivaia. "Advancing Targeted Nanoparticle mRNA Delivery Using Machine Learning". chemrxiv. 2026-10-15T05:37:43.000Z. URL: https://doi.org/10.26434/chemrxiv.15010224

[9] Neiman, G., Costa, M. W., Han, H., Zhao, S., Ng, T., Siemons, B., Nishino, T., Huang, Y., Lal, S., Wu, K., Judge, L., Conklin, B., Srivastava, D., Murthy, N., Healy, K. E.. "A cardiac microphysiological system for screening lipid nanoparticle/mRNA complexes predicts in vivo efficacy for heart transfection". bioRxiv. 2025-03-12T00:00:00.000Z. URL: https://doi.org/10.1101/2025.03.07.642075

[10] Nawaz, M., Tangruksa, B., Heydarkhan-Hagvall, S., Kohl, F., Gonzalez-King Garibotti, H., Jing, Y., Payandeh, Z., Reyahi, A., Jennbacken, K., Wiseman, J., Hultin, L., Lindfors, L., Synnergren, J., Valadi, H.. "Targeted delivery of mRNA to the heart via extracellular vesicles or lipid nanoparticles". bioRxiv. 2025-01-26T00:00:00.000Z. URL: https://doi.org/10.1101/2025.01.25.634881

[11] Sanchez-Fernandez, A., Donnelly, K. A., Bolinsson, H., Borjesdotter, A.-M., Ronnemoes Bobak, T., Erlendsson, S., Mohammadi-Zerankeshi, M., AboulFotouh, K., Kawelah, M. R., Herranz-Trillo, F., Schagerlof, H., Capasso Palmiero, U., Huus, K., Johnston, K. P., Marras, A. E., Nilsson, L.. "Structural heterogeneity in mRNA-LNP subpopulations revealed by AF4-SAXS: implications for cargo loading and cell transfection". bioRxiv. 2026-01-17T00:00:00.000Z. URL: https://doi.org/10.64898/2026.01.16.699683

[12] Islam, F., Das, A., Ashaduzzaman, M., Ding, L., Kumari, N., Dai, R., Oupicky, D.. "Structure-Activity Mapping of Intraperitoneal mRNA-LNPs: Decoupling Tumor and Liver Biodistribution in Pancreatic Cancer". bioRxiv. 2026-03-21T00:00:00.000Z. URL: https://doi.org/10.64898/2026.03.20.712457

[13] Dirvelyte-Valauske, E., Mazerimas, M., Pavliukeviciene, B., Daugelaviciene, N., Kutanovas, S., Kao, C.-Y., Chen, Y.-T., Neniskyte, U., Budvytyte, R.. "Improved Protein Encapsulation and Delivery by Lipid Nanoparticles with Refined Ionizable Lipid Content". bioRxiv. 2026-03-12T00:00:00.000Z. URL: https://doi.org/10.64898/2026.03.10.710763

[14] Worthington, A. K., Borges, B., Echeverri, E. S., Zada, F. M., Lum, T., Davis, A., Jia, K., Perez, B., Graveson, A., Oder, D., Cordero, M., Kim, H., Zenhausern, R., Celik, O., Mccreary, T., Singh, H., Majzoub, R., Shaw, C., Gutierrez-Martinez, P., Omarova, M., Blanchard, C., Burns, S., Cromer, K., Dahlman, J., MacKenzie, T.. "In utero and neonatal lipid nanoparticle delivery achieves robust editing in hematopoietic stem cells.". bioRxiv. 2026-09-22T00:00:00.000Z. URL: https://doi.org/10.1101/2025.05.12.652147

[15] Jackson, C. R., Borsos, M., Appling, N., Jackson, C. R., Coughlin, G. M., Gradinaru, V.. "Defining a Midgestational Window for In Utero Genome Editing of the Fetal Murine Cortex". bioRxiv. 2026-05-01T00:00:00.000Z. URL: https://doi.org/10.64898/2026.04.28.721509

[16] Hamilton, A. G., Thatte, A. S., Xu, J., Luo, Z., Safford, H. C., Swingle, K. L., Muscat-Rivera, J., Kegel, M., Han, X., Joseph, R. A., Murray, A. M., Geisler, H. C., Whitaker, R. C., Xue, L., Spektor, R., Melamed, J. R., Weissman, D., Mitchell, M. J.. "High-throughput in vivo screening using barcoded mRNA identifies lipid nanoparticles with extrahepatic tropism for cancer immunotherapy". bioRxiv. 2025-06-09T00:00:00.000Z. URL: https://doi.org/10.1101/2025.06.09.658602

[17] Ghoshal, B., Chakraborty, D., Nag, M., Varadarajan, R., Jhunjhunwala, S.. "Ex Vivo Delivery of mRNA to Immune Cells Via a Non-Endosomal Route Obviates the Need for Nucleoside Modification". bioRxiv. 2024-11-03T00:00:00.000Z. URL: https://doi.org/10.1101/2024.10.31.621287

[18] Xiao, Y., Zhou, S., Zhu, T., Du, Y., Hu, Q., Guo, X., Wang, L., Deng, S., Ren, K.. "Phase-separating enzyme-responsive condensates for inhalable delivery of biomacromolecules". bioRxiv. 2025-06-23T00:00:00.000Z. URL: https://doi.org/10.1101/2025.06.17.660252

[19] Song, P., Su, J., Vraa, C. L. K., Gockert, M., Fjelstrup, S., Hager, H., Kjems, J.. "Optimizing C14120-based LNPs for in vitro and in vivo mRNA delivery". bioRxiv. 2025-09-06T00:00:00.000Z. URL: https://doi.org/10.1101/2025.09.02.673705

[20] Lee, Y., Jeong, H., Kim, E., Hwang, Y., Byeon, Y., Kang, H., Choi, M. S., Jeong, E. H., Kwak, J. H., Kang, M.-S., Kim, O.-H., Eom, S., Ahn, J. H., Lee, Y. J., Byeon, S. H., Kim, S.-J., Lee, J., Lee, H.. "A Rapidly Excretable, ROS-Scavenging Ionizable Lipid Decouples mRNA Delivery Potency from Toxicity". bioRxiv. 2026-04-09T00:00:00.000Z. URL: https://doi.org/10.64898/2026.04.07.716828

How this survey was made

This survey answers one question: “How have lipid nanoparticle formulations been optimized for delivering CRISPR/Cas9 ribonucleoproteins versus mRNA cargos to embryonic tissues in vivo, and what biodistribution and toxicity tradeoffs distinguish these cargo types?”

We keep our own collection of recent preprints and refresh it every day from bioRxiv, medRxiv, arXiv, PsyArXiv, SocArXiv and ChemRxiv. For each survey, we search that collection for recent work and keep the strongest matches: papers that use the same terms as the question, and papers on the same topic in different words.

The text was written from the titles and abstracts of those papers; the full texts were not read. Statements that draw on a particular paper are marked with a number in brackets. The reference list just above is the 20 papers whose work this text draws on, chosen from the 128 papers those searches returned; the rest of the shortlist is not printed. It is not a complete review of everything published on the question.

Treat this as a starting point and a reading list, not a peer-reviewed review. The searches behind this survey ran on October 22, 2026 and covered work from the previous two years.