Formulation & Performance — Lipid R6

Research application and evidence scope

R6 was selected from the AGILE candidates for RAW 264.7 macrophage transfection. Its activity is cell-type dependent and is not evidence of universal superiority or in vivo macrophage targeting. H9 is the separate lead for the muscle-delivery experiments.

Values shown are literature-reported results for the cited formulation and study conditions. They do not guarantee performance of the supplied product in other formulations or experiments.

Formulation records and source discrepancy

The main-text optimized ratio and SI top run #2 disagree. Both are retained as separate source records. The manuscript does not resolve which composition maps to every later performance panel. The normalized tables below are arithmetic illustrations, not a corrected or validated preparation recipe.

Source: Main Methods, “LNP formulation and characterization”; SI p.19, Table 4, run #2.

Component molecular weights and mass tables

Calculated basis: 10 µmol total lipid components. Mass = component amount × MW / 1000 in mg. DOTAP uses the catalog chloride-salt MW; DMG-PEG2000 uses 2507. The SI table labels its PEG component only as “PEG”; using DMG-PEG2000 for that table is a mapping to the main Methods, not an independent identity measurement.

Pre-optimization composition

Methods optimized ratio — unresolved

SI Table 4 run #2 — unresolved

For unformulated R6, 1 mg corresponds to 1.8116 µmol, and a 1 mM stock requires 0.55199 mg/mL. These chemical stock conversions do not establish solvent compatibility or an RNA formulation protocol.

Source: Reported composition: main Methods and SI Table 4. Normalized fractions and masses: calculated from catalog MW inputs.

Reported preparation and analytical methods

Mix the aqueous mRNA phase in 10 mM citrate buffer with the ethanol lipid phase at 3:1 v/v. The initial screening used an automated liquid handler, crude lipids and direct cell testing. Subsequent experiments used manual pipetting and dialysis against 1× PBS in a 20 kDa cutoff cassette for 6 h at 4 °C. The cited Methods do not state citrate pH, a microfluidic flow rate or an N/P value. The general 10:1 ratio refers to ionizable lipid:mRNA mass, not N/P.

Particle size, PDI and zeta potential were measured by light scattering/electrophoretic methods; RNA encapsulation used a RiboGreen assay. Apparent LNP pKa was determined from the half-maximum TNS fluorescence point over pH 2–11. An exact R6 pKa, particle size, PDI or encapsulation percentage is not assigned from the available R6-specific records.

Source: Main Methods, “LNP formulation and characterization”.

GFP mRNA delivery to RAW 264.7 macrophages

These are the percentages displayed on the representative flow-cytometry distributions, not means with replicate uncertainty. From those displayed percentages, R6/MC3 is 5.20-fold and R6/H9 is 7.54-fold (calculated); no significant superiority over ALC-0315 is inferred from 51.0% versus 48.1%. Imaging was reported with n = 3; replicate GFP percentages are not provided in the downloaded Source Data sheet.

Source: Main Fig. 4f–g and Methods, “Transfection test in RAW 264.7 cells”.

Cell-type-specific luciferase expression

mTP is log₂(mean luminescence in treated cells / mean luminescence in untreated cells) at 24 h. It is not a transfected-cell percentage. General cell screening used 0.1 µg Fluc mRNA per well in 96-well plates.

The values above are recalculated from the six source-data entries per group in Fig. 4e. R6 leads in RAW 264.7, whereas H9 leads in HeLa. The paper describes the R6/H9 comparison as composition-matched, but later Methods and SI optimized compositions differ; do not assign a single unverified optimized composition to this panel.

Source: Main Fig. 4e; Methods, “In vitro high-throughput screening”; Source Data, “Figure 4”, rows 9–11.

DoE improvement in RAW 264.7 cells

Calculated post/pre mean ratio: 1.53-fold. The source reports p = 0.0353. This RLU comparison is separate from the GFP-positive-cell endpoint and the Log2 DoE response.

Source: SI p.45 Fig. S26a; Source Data, “Supplementary Figure 26”, B2:C7.

Source: SI p.19, Table 4.

In vivo boundaries and vaccine results

Source Data S28b contains four titers per group: R6 400, 200, 800, 800; ALC-0315 51,200, 102,400, 51,200, 25,600. Their arithmetic means are 550 and 57,600 (calculated), with source p = 0.0021 for R6 versus ALC-0315. The SI caption includes n = 3, so the four-entry antibody dataset is reported separately rather than silently changing the replicate count. R6 is therefore presented for macrophage transfection research, not as the best vaccine lipid in this study.

Source: SI pp.46–47, Figs. S27–S28; Source Data, “Supplementary Figure 28”.

Structure and synthesis evidence

The assigned structure is 2-((3-(diethylamino)propyl)amino)-N-undecylheptadecanamide, a neutral amino-amide with one tertiary amine, one secondary amine and one amide nitrogen. The stereochemistry at the substituted carbon is not specified. The explicitly named R6 synthesis combines N,N-diethylpropane-1,3-diamine, palmitaldehyde and 1-isocyanoundecane through a three-component Ugi reaction; purified R6 was reported as a colorless oil in 37.56% yield. The verified drawing and reagent mapping agree with C35H73N3O and calculated average MW 551.99.

The SI reports [M+H]⁺ calculated 552.58 and observed ESI m/z 552.73. That is not an average molecular weight or high-resolution mass confirmation. The source uses the neutral formula alongside the protonated-ion label; the actual protonated ion is C35H74N3O⁺. The main-text narrative gives tail-length descriptions that differ from the explicit R6 synthesis and drawing, so the structure assignment here follows the named synthesis and Fig. S24.

Source: SI p.8 lines 296–306; p.11 R6 mass spectrum; p.14 R6 ¹H NMR; p.43 Fig. S24.

References

Xu, Y.; Ma, S.; Cui, H.; et al. AGILE platform: a deep learning powered approach to accelerate LNP development for mRNA delivery. Nature Communications 2024, 15, 6305. DOI: 10.1038/s41467-024-50619-z.

Supplementary Information, pp. 1–57 · Source Data workbook.

Study content is paraphrased and calculations are identified; source article licensed under CC BY 4.0.