Nanobody-Decorated Lipid Nanoparticles for Enhanced mRNA Delivery to Tumors In Vivo

Abstract

Prostate cancer (PCa) ranks as the fifth leading cause of cancer-related deaths among men worldwide. In 10-20% of the cases, PCa progresses to an incurable, castration-resistant stage. Castration-resistant PCa cells often overexpress prostate-specific membrane antigen (PSMA), a membrane protein that may serve as their Achilles' heel. Over the past decades, RNA-based therapeutics have emerged as promising treatments for a vast array of diseases, including cancer. In this study, with the ultimate goal of developing a targeted therapy for PCa, lipid nanoparticles (LNPs) are decorated with an anti-PSMA nanobody using click chemistry with a PEG-lipid. Direct stochastic optical reconstruction microscopy (dSTORM) and cluster analysis confirm the presence of at least one nanobody on the surface of 80% of LNPs. These anti-PSMA LNPs exhibit enhanced and specific uptake, and mRNA transfection in PSMA+ cancer cells both in vitro and in a Zebrafish (ZF) metastatic PCa xenograft model. Additionally, in a mouse PSMA-positive xenograft model, systemic administration results in increased LNP accumulation, but not functional mRNA delivery. These findings underscore both the potential and the challenges of using a PSMA-targeted lipid nanoparticle system for mRNA delivery into advanced prostate cancer tumors.

Keywords: mRNA‐lipid nanoparticles; nanobody; prostate cancer; prostate specific membrane antigen (PSMA); targeted delivery.

Figure 1

Schematic overview illustrating click chemistry and the post‐insertion of VHHs onto the surface of LNPs. A) Sortase A cuts LPETG (sortase A tag) to allow coupling of an azide peptide to the C‐terminal of VHH. B) Click chemistry reaction between Azide‐VHH and DBCO‐PEG‐lipid. C) Post‐insertion of DSPE‐PEG‐VHH onto the surface of LNPs followed by dialysis overnight to obtain VHH‐decorated mRNA‐LNPs.

Figure 2

Through dSTORM imaging and cluster analysis we identified VHH on the surface of anti‐PSMA LNPs. LNPs were labeled with Cy5‐DSPE (pink), and VHH was detected after incubation with 10 µg mL⁻¹ anti‐Myc‐Atto 488 antibody (cyan). A) ONI Super‐resolution microscopy image of a frame with clusters for anti‐PSMA LNPs (left) and nontargeted LNPs (NT LNPs, right). B) Through clustering analysis, we determined that ≈80% of anti‐PSMA LNPs contained at least one VHH on their surface. TIRF acquisition was performed with a Nanoimager‐S (ONI) and CODI online analysis tool was used for the filtering of localizations and HDBSCAN cluster analysis. Data represent mean ± SD (n = 2 technical replicates).

Figure 3

LNP uptake and mRNA transfection of anti‐PSMA LNPs in PSMA⁺ cells are receptor‐mediated. A,B) Flow cytometry analysis for uptake (A) and mRNA transfection (B) 24 h after addition of 200 ng of targeted‐LNPs (0.2% DSPE‐PEG(2000)‐VHH post‐insertion) encapsulating Cy5‐labeled EGFP mRNA in B16‐F10‐PSMA & B16‐F10 cell lines. In parallel, B16‐F10‐PSMA & B16‐F10 cells were subjected to 5‐min incubations with 15, 1.5, or 0 nmols of unconjugated anti‐PSMA VHH. After the VHH incubation, we added 200 ng of anti‐PSMA or R2‐LNPs (0.2% DSPE‐PEG(2000)‐VHH post‐insertion) encapsulating Cy5‐labeled EGFP mRNA to each well. 24 h after LNP addition, we measured the LNP uptake and mRNA transfection results by flow cytometry. C,D) Results are shown as a proportion of the LNP uptake (C) and mRNA transfection (D) for each of the LNP treatments in respect to their paired conditions without pre‐treatment of unconjugated anti‐PSMA‐VHH. A Two‐Way ANOVA with Šídák's correction for multiple comparisons test was performed comparing the mean signals of each condition in between the different cell lines. ****, p‐value < 0.0001; ***, p‐value < 0.001; **, p‐value < 0.01; *, p‐value < 0.05; ns: no significant difference. Data represent mean ± SD (n = 3 wells) with at least 5000 cells per well.

Figure 4

Anti‐PSMA LNPs show targeted mRNA delivery to metastatic PCa cells in a Zebrafish Xenograft model. A) Time frame of in vivo experiments in ZF Casper embryos. LNCaP (PSMA⁺) cells were engrafted in the PVS at 2 dpf. LNPs were i.v. injected in the Duct of Cuvier at 3 dpf. After 7 h, embryos were fixed with 4% PFA in egg‐water. Imaging was performed using a Leica SP5 confocal microscope. Color legend: Green = LNCaP cells marked with Membright‐488. Red = LNPs containing 0.2% DSPE‐Rhodamine. Blue = BFP protein expression. Injected dose: ≈1 mg kg⁻¹ mRNA. Injection volume: 2 nL. PVS = perivitelline space; i.v. = intravenous; ZF = zebrafish; dpf = day post fertilization. B,C) ZF caudal biodistribution of anti‐PSMA LNPs (B, red) or R2‐LNPs (C, red) and respective merge with LNCaP cells (green) and BFP expressing cells (blue) in the tail. D,E) Pearson's correlation coefficients (r) of LNCaP cells (green) over LNPs (red) signals (D), and LNCaP cells (green) over BFP (blue) signals (E) comparing anti‐PSMA LNPs vs R2‐LNPs conditions, respectively. Anti‐PSMA LNPs showed enhanced targeting (B,D) and functional BFP mRNA transfection in LNCaP cells (B,E) when compared to R2‐LNPs (C,D and C,E). Pearson's colocalization coefficients (r) between the different channels were calculated on Fiji‐ Image J software. Significance of the differences between groups was determined by a two‐tailed unpaired t‐test, for datasets previously tested with a normal (Gaussian) distribution by Shapiro–Wilk test. Data represent mean ± SD of n = 10 (D), n = 6 (E) embryos per condition. ****, p‐value < 0.0001; ***, p‐value < 0.001.

Figure 5

Anti‐PSMA LNPs show enhanced tumor accumulation but not functional mRNA transfection in mouse tumors after 24 h. A,B) Measurement of Cy5 signal in organ tissue lysates (B) and in tumor lysates alone. C,D) Measurement of luciferase signal in organ tissue lysates (D) and in tumor lysates alone. Biodistribution of LNPs by quantification of Cy5 signal (A,B) and functional mRNA transfection by quantification of luciferase signal (C,D) are expressed as percentage of the injected dose per gram (% ID/g) of tissue. E,F) Quantification of Cy5 signal from tumor sections of animals treated with anti‐PSMA LNPs, R2 LNPs or PBS determined by Cy5 raw intensity (E) and % of Cy5⁺ area (F). For the graphs (A,C) a two‐way ANOVA with Šidák correction for multiple comparisons test was performed comparing the mean signals from each organ between the treatment groups. For the graphs (B,D) statistical differences were assessed with a two‐tailed unpaired t‐test. For the graphs (E,F) statistical differences were assessed with a One Way Anova nonparametric Kruskal–Wallis test with Dunn's multiple comparisons. Data represent mean ± SD (n = 4 animals) for graphs A–D and n = 3‐4 animals for graphs E and F. ****, p‐value < 0.0001; ***, p‐value < 0.001; **, p‐value < 0.01; *, p‐value < 0.05; ns: no significant difference.