Exosome-mediated Delivery of Hydrophobically Modified siRNA for Huntingtin mRNA Silencing

Abstract

Delivery represents a significant barrier to the clinical advancement of oligonucleotide therapeutics for the treatment of neurological disorders, such as Huntington's disease. Small, endogenous vesicles known as exosomes have the potential to act as oligonucleotide delivery vehicles, but robust and scalable methods for loading RNA therapeutic cargo into exosomes are lacking. Here, we show that hydrophobically modified small interfering RNAs (hsiRNAs) efficiently load into exosomes upon co-incubation, without altering vesicle size distribution or integrity. Exosomes loaded with hsiRNAs targeting Huntingtin mRNA were efficiently internalized by mouse primary cortical neurons and promoted dose-dependent silencing of Huntingtin mRNA and protein. Unilateral infusion of hsiRNA-loaded exosomes, but not hsiRNAs alone, into mouse striatum resulted in bilateral oligonucleotide distribution and statistically significant bilateral silencing of up to 35% of Huntingtin mRNA. The broad distribution and efficacy of hsiRNA-loaded exosomes delivered to brain is expected to advance the development of therapies for the treatment of Huntington's disease and other neurodegenerative disorders.

Figure 1

Efficient loading of exosomes with hsiRNAs. (a) hsiRNA^HTT schematic and PyMOL model. (b) Flowchart of exosome loading procedure: co-incubation of Cy3-hsiRNA and exosomes results in Cy3-hsiRNA-loaded exosomes that pellet by ultracentrifugation. (c) Bar graphs showing the percent of Cy3-hsiRNA or Cy3-hsiRNA-exosomes in the pellet (black) and supernatant (white) after ultracentrifugation (n = 3; mean ± SD). (d) hsiRNA loading does not affect exosome size distribution (nonloaded exosomes, solid line; hsiRNA-loaded exosomes, dashed line). (e) Electron microscopy of exosomes in the absence (left) or presence (right) of hsiRNA. Exosomes remain intact when loaded with hsiRNA (scale bar = 100 nm). hsiRNA, hydrophobically modified small interfering RNA.

Figure 2

Internalization of hydrophobic Cy3-hsiRNA-loaded exosomes by primary cortical neurons. (a) Kinetics of Cy3-hsiRNA^HTT-loaded exosomes shows significant co-localization between exosomes and hsiRNAs. hsiRNA^HTT labeled with Cy3 (red); exosomes labeled with PKH67 (green); nuclei labeled with Hoechst (blue). Samples were visualized at 63× magnification. Scale bar = 10 µm. Arrows show hisRNA^HTT-exosomes. (b) Bar graphs showing the absolute quantification of Cy3-fluorescence in cells upon incubation with Cy3-hsiRNA (black) or Cy3-hsiRNA-exosomes (white) (n = 5; mean ± SD; *P < 0.05, **P < 0.01, paired t-test comparing the Cy3 intensity of hsiRNA nonloaded versus loaded in exosomes). (c) Quantification of Cy3-hsiRNA^HTT taken up by primary neurons with increasing amounts of Cy3-hsiRNA^HTT-loaded exosomes. The level of Cy3-labeled hsiRNAs was measured by high-performance liquid chromatography (n = 3 technical replicates; mean ± SD). hsiRNA, hydrophobically modified small interfering RNA.

Figure 3

Concentration-dependent silencing of Huntingtin mRNA and protein by hsiRNA^HTT-loaded exosomes in primary cortical neurons. (a) Graph of Htt mRNA levels in cells treated with increasing amounts of hsiRNA^HTT-loaded exosomes (solid line), hsiRNA^NTC-loaded exosomes (dotted line), or nonloaded exosomes (dashed line) for 7 days. Htt mRNA levels are normalized to Ppib mRNA (housekeeping control) and presented as percent of untreated control (n = 4 biological replicates, each performed with n = 3 technical replicates; mean ± SD; *P < 0.05, **P < 0.01, paired t-test comparing each concentration to hsiRNA^NTC-loaded exosomes). (b) Graph of HTT protein levels in cells treated with increasing amounts of hsiRNA^HTT-loaded exosomes (solid line; n = 5 biological replicates) or hsiRNA^NTC-loaded exosomes (dotted line; n = 3 biological replicates) for 7 days. HTT signal was normalized to Tubulin. Data at each concentration are presented as percent of untreated control samples (mean ± SD; *P < 0.05, **P < 0.01, paired t-test comparing each concentration to untreated samples). hsiRNA, hydrophobically modified small interfering RNA; NTC, nontargeting control.

Figure 4

Exosomes promote bilateral distribution of hsiRNAs. (a,b) Images of ipsilateral (left) and contralateral (right) brain sections after infusion of Cy3-hsiRNA^HTT-loaded exosomes (red) (a) or unlabeled hsiRNA^HTT-loaded exosomes (b). Nuclei stained with Hoechst (blue). Samples were visualized at 2.5× magnification. Scale bar = 1 mm. All pictures were acquired with the same microscope settings. (c,d) Plot showing tissue concentrations of hsiRNA^HTT (c) or hsiRNA^sFlt1 (d) guide strands measured by peptide nucleic acid hybridization in biopsies from the ipsilateral (IL) and contralateral (CL) striata of mice treated with hsiRNA-loaded exosomes, hsiRNAs alone, or artificial cerebrospinal fluid (CSF) (hsiRNA^HTT, n = 7 mice per group; hsiRNA^sFlt1, n = 9 mice per group). Each point in the plot represents the level of hsiRNA guide strands in an individual mouse, and the mean ± SD of each group is indicated. Data were analyzed using a nonparametric Kruskal–Wallis test, and the mean of each column was compared to the mean of the corresponding artificial CSF control column. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. hsiRNA, hydrophobically modified small interfering RNA.

Figure 5

hsiRNA^HTT-loaded exosomes induce bilateral Htt mRNA silencing in vivo in mouse brain. (a) Plots showing Htt mRNA levels in biopsies from the ipsilateral (closed) and contralateral (open) striata of mice treated with artificial cerebrospinal fluid (CSF), unloaded exosomes, hsiRNA^NTC-loaded exosomes, 0.5 and 1 µg/day hsiRNA^HTT, or 0.5 and 1 µg/day hsiRNA^HTT-loaded exosomes (red). Exosomes were delivered at 2–4 × 10¹⁰ particles/day. Level of Htt mRNA was normalized to Ppib mRNA (housekeeping control), and data are presented as percent of untreated control (n = 6–10 mice per group). Each point represents the averaged level of Htt mRNA from three biopsies of an individual mouse, and the mean ± SD of each group is indicated. (b) Bonferroni's multiple comparison tests between groups treated with artificial CSF, exosomes, hsiRNA^HTT, or hsiRNA^HTT-exosomes and group treated with hsiRNA^NTC-loaded exosomes. IL, ipsilateral; CL, contralateral; hsiRNA, hydrophobically modified small interfering RNA; NTC, nontargeting control.