A Polymeric Nanoparticle Formulation for Targeted mRNA Delivery to Fibroblasts

Abstract

Messenger RNA (mRNA)-based therapies offer enhanced control over the production of therapeutic proteins for many diseases. Their clinical implementation warrants formulations capable of delivering them safely and effectively to target sites. Owing to their chemical versatility, polymeric nanoparticles can be designed by combinatorial synthesis of different ionizable, cationic, and aromatic moieties to modulate cell targeting, using inexpensive formulation steps. Herein, 152 formulations are evaluated by high-throughput screening using a reporter fibroblast model sensitive to functional delivery of mRNA encoding Cre recombinase. Using in vitro and in vivo models, a polymeric nanoformulation based on the combination of 3 specific monomers is identified to transfect fibroblasts much more effectively than other cell types populating the skin, with superior performance than lipid-based transfection agents in the delivery of Cas9 mRNA and guide RNA. This tropism can be explained by receptor-mediated endocytosis, involving CD26 and FAP, which are overexpressed in profibrotic fibroblasts. Structure-activity analysis reveals that efficient mRNA delivery required the combination of high buffering capacity and low mRNA binding affinity for rapid release upon endosomal escape. These results highlight the use of high-throughput screening to rapidly identify chemical features towards the design of highly efficient mRNA delivery systems targeting fibrotic diseases.

Keywords: CRISPR/Cas9; gene edition; high-throughput screening; messenger RNA; polymeric nanoparticles.

Figure 1

High‐throughput screening of polymer library for mRNA delivery. a) NPs were generated after complexation of mRNA with polymers constituted by a fixed diacrylate moiety (P1) with varying bisacrylamides (A–E) and amines (1–32). Polymer library was screened for the delivery of mRNA encoding Cre recombinase in mouse embryonic fibroblast reporter cell line, expressed by functional activation of GFP at 48 h post‐transfection. b) Selected polymers (highlighted in blue after exclusion of false positives) showed comparable transfection efficiency to Lipofectamine 2000. Non‐cytotoxic polymers inducing less Cre‐mediated recombination than naked Cre mRNA were excluded from this analysis. Data are expressed as mean ± SEM (n = 3–14). c) Chemical structures of common bisacrylamides and amines in identified hits. d) Transfection efficiency of NPs containing mRNA Cre recombinase after polymer purification by dialysis. (d.1) Cre‐mediated recombination was quantified by high‐content imaging. Data are expressed as mean ± SEM (n = 6–17). (d.2) Representative fluorescence microscope images of lead NP candidate (P1E28) compared to Lipofectamine 2000. Scale bars = 100 µm. Data in (b) and (d.1) were analyzed by one‐way ANOVA with post hoc Dunnett's multiple comparisons test against free mRNA and untreated controls, respectively: (*), p < 0.05; (**), p < 0.01; (***), p < 0.001; (****), p < 0.0001.

Figure 2

Hit validation and cellular tropism. a) Identified polymers were used to transfect GFP mRNA in human dermal fibroblasts. (a.1) Transfection efficiency and (a.2) representative microscope images of human dermal fibroblasts transfected with GFP mRNA. Scale bars = 100 µm. b) Transfection of GFP mRNA in human endothelial cells, c) human keratinocytes, and d) human monocytes. Data are expressed as mean ± SEM (n = 3) except for monocytes (n = 6). One‐way ANOVA with post hoc Tukey's multiple comparisons test was performed: (ns), p > 0.05; (*), p < 0.05; (***), p < 0.001; (****), p < 0.0001.

Figure 3

Development of an RNA‐based CRISPR/Cas9 delivery system. a) CRISPR/Cas9 is based on the delivery of Cas9 (as mRNA) alongside a guide RNA (sgRNA). Gene edition was performed on mouse embryonic fibroblast reporter cell line after Cre‐mediated recombination to promote stable GFP expression. Gene edition resulted in decreased GFP expression in transfected cells. b) Representative images of GFP knockout induced by P1E28 alongside respective controls, obtained at day 3 post‐transfection using high‐content imaging. Scale bars = 100 µm. c) Long‐term effects of gene edition quantified at day 10 post‐transfection using flow cytometry. (c.1) Representative flow cytometry scatter plots of cells treated with P1E28 formulation alongside controls demonstrate GFP knockout. (c.2) Percentage of GFP knockout normalized to the initial GFP expression, assessed by flow cytometry. GFP knockout was calculated after normalizing the number of GFP negative cells in each treatment compared to the untreated control. Data are expressed as mean ± SEM (n = 3–6). One‐way ANOVA with post hoc Dunnett's multiple comparisons test was performed: (*), p < 0.05. d) Cellular uptake of NPs complexed with Cas9 mRNA and a fluorescently labeled sgRNA (ATTO550) by dermal fibroblasts was assessed by flow cytometry 4 h after transfection. (d.1) Percentage of cells internalizing NPs with ATTO550‐labelled sgRNA and (d.2) representative scatter plots of cells treated with P1E28 or DharmaFECT Duo. Data are expressed as mean ± SEM (n = 3). One‐way ANOVA with post hoc Tukey's multiple comparisons test was performed: (**), p < 0.01. e) Endosomal disruption induced by P1E28 was measured by quantifying the formation of galectin‐9 foci. (e.1) Representative confocal microscopy images of reporter fibroblasts 4 h after incubation with P1E28 or DharmaFECT Duo. Hydroxychloroquine (HCQ) was used as a positive control (60 µm). Scale bars = 20 µm (insets = 6.5 µm). (e.2) Number of galectin‐9 foci per cell. Data are expressed as mean ± SEM (n = 2–3 independent experiments, with 3–4 confocal images acquired for each replicate). One‐way ANOVA with post hoc Tukey's multiple comparisons test was performed: (*), p < 0.05; (**), p < 0.01.

Figure 4

In vivo validation of targeted mRNA delivery to skin fibroblasts mediated by P1E28 during wound healing. a) R26‐tdTomato mice were intradermally injected with Cre mRNA delivered by P1E28 or Lipofectamine 2000. Skin samples were obtained from wounds at day 7 post‐injection. (a.1) Representative fluorescence microscope images of skin sections illustrate the transfection efficiency of P1E28, generating tdTomato⁺ cells (red) after Cre‐mediated recombination. Scale bars = 500 µm. (a.2) Quantification of generated tdTomato⁺ cells normalized by tissue section area. Data are expressed as mean ± SEM (n = 6 animals, 2 wounds per animal). Unpaired two‐sample t‐test was performed: (*), p < 0.05. b) Tropism of P1E28 polyplexes was characterized by immunohistochemical staining of skin sections containing tdTomato+ cells. (b.1) Representative fluorescence microscope images illustrate that the vast majority of α‐SMA+ activated fibroblasts (light blue) co‐localized with tdTomato fluorescent areas (red). Scale bars = 500 µm. (b.2) Co‐localization of each cell type with tdTomato was obtained after calculating the Manders’ overlap coefficient. Data are expressed as mean ± SEM (n = 6 animals, 3–5 images per animal). One‐way ANOVA with post hoc Tukey's multiple comparisons test was performed: (****), p < 0.0001.

Figure 5

Structure‐activity relationships of polymeric NPs for mRNA delivery. a) Physicochemical properties of the identified hits: (a.1) surface charge, (a.2) hydrodynamic diameter, and (a.3) complexation efficiency. Results are expressed as mean ± SEM (n = 3). In (a.1) and (a.2), data were analyzed by one‐way ANOVA with post hoc Tukey's multiple comparisons test: (*), p < 0.05; (**), p < 0.01. In (a.3), data were analyzed by two‐way ANOVA with post hoc Sidak's multiple comparisons test: (*), p < 0.05. b) Heparin replacement assay estimated the polymers’ binding affinity to Cre mRNA after quantification of fluorescence intensity of bands in each lane corresponding to a different heparin dose. Data represent the mean value of 2–3 gel replicates. Horizontal dashed line corresponds to 50% mRNA release. c) Acid‐base titration for the selected polymers dissolved in 150 mm NaCl, adjusted for pH 3 with HCl. d) Multiple linear regression of polymer buffering capacity and mRNA binding affinity enabled the correlation of these parameters with transfection efficiency. (d.1) Standard least squares effect leverage with two‐way interactions described the significant interaction of these parameters in transfection efficiency, with negative coefficient Heparin*Buffer indicating an inversely proportional relationship. (d.2) Heatmap representation of the predicted model with highlighted experimental points corresponding to selected polymers.

Figure 6

Molecular mechanism of P1E28 formulation for fibroblast targeting. a) Size of polyplexes generated by complexation of P1E28 with Cre mRNA. (a.1) Representative TEM image and (a.2) NP diameter distribution show the formation of small NPs (87% of measured NPs < 200 nm). Scale bar = 500 nm. b) Recombination in mouse reporter fibroblast cell model after endocytosis inhibition of P1E28 via clathrin‐ (chlorpromazine) and caveolin‐mediated endocytosis (filipin III), as well as actin‐related processes (cytochalasin D). Data are expressed as mean ± SEM (n = 5–6). One‐way ANOVA with post hoc Dunnett's multiple comparisons test was performed: (**), p < 0.01; (***), p < 0.001; (****), p < 0.0001. c) Transfection efficiency was determined by quantifying the number of cells expressing GFP, after delivery of GFP mRNA to human dermal fibroblasts with or without receptor blocking using monoclonal antibodies targeting CD26 and FAP. Data are expressed as mean ± SEM (n = 3). Two‐way ANOVA with post hoc Sidak's multiple comparisons test was performed: (*), p < 0.05; (**), p < 0.01. d) Impact of receptor blocking on cellular uptake of NPs complexed with Cas9 mRNA and a fluorescently labeled sgRNA (ATTO550) was assessed by flow cytometry. (d.1) Percentage of cells internalizing NPs with ATTO550‐labelled sgRNA 4 h after transfection and (d.2) representative scatter plots of cells treated with P1E28 with or without antibody blocking. Data are expressed as mean ± SEM (n = 3). Two‐way ANOVA with post hoc Sidak's multiple comparisons test was performed: (**), p < 0.01; (****), p < 0.0001.