Enhancing Gene Delivery to Breast Cancer with Highly Efficient siRNA Loading and pH-Responsive Small Extracellular Vesicles

Abstract

Small extracellular vesicles (sEVs) are promising nanocarriers for drug delivery to treat a wide range of diseases due to their natural origin and innate homing properties. However, suboptimal therapeutic effects, attributed to ineffective targeting, limited lysosomal escape, and insufficient delivery, remain challenges in effectively delivering therapeutic cargo. Despite advances in sEV-based drug delivery systems, conventional approaches need improvement to address low drug-loading efficiency and to develop surface functionalization techniques for precise targeting of cells of interest, all while preserving the membrane integrity of sEVs. We report an enhanced gene delivery system using multifunctional sEVs for highly efficient siRNA loading and delivery. The integration of chiral graphene quantum dots enhanced the loading capacity while preserving the structural integrity of the sEVs. Additionally, lysosomal escape is facilitated by functionalizing sEVs with pH-responsive peptides, fully harnessing the inherent homing effect of sEVs for targeted and precise delivery. These sEVs achieved a 1.74-fold increase in cytosolic cargo delivery compared to unmodified sEVs, resulting in substantial gene silencing of around 73%. Our approach has significant potential to advance sEV-based gene delivery in order to accelerate clinical progress.

Keywords: GALA; gene delivery; graphene quantum dots; lipid nanoparticles; lysosomal escape.

Figure 1.

Schematic illustration of siR-DG/GALA-sEVs. (a) Preparation of siR-DG/GALA-sEV and surface charge conversion in response to pH changes. (b) Mechanism of lysosomal escape for siR-DG/GALA-sEVs to achieve cytoplasmic siRNA delivery.

Figure 2.

Constituents for multifunctional sEVs. Characterization of MCF-7 derived sEVs: (a) Spherical nanostructure analysis by transmission electron microscope (TEM). (b) Size distribution and particle number of sEV analysis by nanoparticle tracking analysis (NTA). (c) sEV marker analysis by Western blot. (d) Surface charge analysis by a Zetasizer (n = 4, mean ± s.e.). One-way ANOVA with Tukey’s post-test. ns, not significant. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. Characterization of chiral graphene quantum dots (GQDs): (e) Nanostructure analysis by TEM. (f) Size distribution based on TEM images (n = 100, mean ± s.d.). (g) Chiroptical property analysis by circular dichroism (CD). (h) Fluorescence intensity (FI) spectra. (i) Chemical composition analysis by Fourier transform infrared (FTIR) spectra. (j) Chemical structures of pH-responsive GALA and cholesterol-conjugated GALA.

Figure 3.

Optimization of multifunctional sEVs. (a) Confocal laser scanning microscopy (CLSM) images of $4\mu \mathrm{M}$ FAM-GALA-chol functionalized sEVs (10⁹ particles/mL). Each CLSM channel represents the following: green for FAM-GALA; red for PKH26-labeled sEV membrane. (b) Fluorescence recovery of FAM-labeled on GALA-chol before and after sEV membrane lysis (n = 4, mean ± s.d.). Excitation: 490 nm; Emission: 520 nm. (c) CLSM images of $15\mu \mathrm{M}$ D-GQDs loaded sEVs (10⁹ particles/mL). CLSM channel represents the following: blue for D-GQDs as sEV cargo. (d) Permeation efficiency quantified by counting D-GQDs loaded sEVs in CLSM images over the total number of sEVs measured by NTA (n = 4, mean ± s.d.). (e) Charge change comparison of sEVs, GALA-sEVs, and DG/GALA-sEVs in pH 4–7 (n = 4, mean ± s.e.). (f) CLSM images of $10\mu \mathrm{M}$ Cy5-siRNA loaded sEVs (3.33 × 10⁸ particles/mL). CLSM channel represents the following: red for Cy5-labeled siRNA. (g) Permeation efficiency quantified by counting Cy5-siRNA loaded sEVs in CLSM images over the total number of sEVs measured by NTA (n = 4, mean ± s.d.). (h) Loading efficiency quantification based on loss efficiency from flow-through eluent analysis using agarose gel (n = 3, mean ± s.d.). Structural integrity of DG/GALA-sEVs confirmed with analysis using (i) TEM and (j) NTA. Arrows indicate D-GQDs within GALA-sEVs. (k) Agarose gel analysis of siRNA-loaded sEVs after incubation with FBS for 0, 1, 3, and 6 h. (l) Agarose gel analysis of washed-out eluent after siRNA loading into sEVs. One-way ANOVA with Tukey’s post-test. ns, not significant. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 4.

Cellular uptake and lysosomal escape of GALA-sEVs. $8\mu \mathrm{M}$ GALA-chol per 10⁹ sEVs. (a) Cytoplasmic cargo delivery analysis of DG/GALA-sEVs. Each CLSM channel represents: red for the cytosol of MCF-7 cells; blue for D-GQDs as sEV cargo. (b) Quantification of cytoplasmic delivery of DG/GALA-sEV cargo in GALA-chol concentration variants onto 10⁹ sEVs (n = 3, mean ± s.d.). (c) Time-dependent lysosomal escape analysis of DG/GALA-sEVs. Each CLSM channel represents: red for lysosomes within MCF-7 cells; blue for D-GQDs as sEV cargo. Arrows indicate lysosomes without D-GQDs postescape. The quantification (n = 3, mean ± s.d.) of (d) Pearson’s correlation coefficient (PCC) between sEV cargo (D-GQDs; blue) and lysosomes (red), (e) sEV cargo area per cell, (f) sEV cargo FI per cell. (g) siRNA release analysis after cytoplasmic delivery of siR-DG/GALA-sEVs. Each CLSM channel represents: red for Cy5-labeled siRNA loaded within sEVs; blue for D-GQDs as sEV cargo; green for cellular membrane. One-way ANOVA with Tukey’s post-test. ns: not significant. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 5.

Successful siRNA transfection via siR-DG/GALA-sEVs and doxorubicin (Dox) combination treatments. (a) Schematic illustration of TGF-β silencing and following ECM change. Western blot analysis (n = 3, mean ± s.d.) of TGF-β1 expression levels in MCF-7 cells after siRNA transfection for the following: (b, d) Dose-dependent test of siR-DG/GALA-sEVs. (c, e) Comparison of gene silencing effect with competitive samples (Ctrl, siRNA, siR-DG, siR-DG/sEV, siR-DG/GALA-sEV, siR-Lipo/sEV). 0.5k sEVs per cell. Cell migration test: (f) Scratch assay monitored at 0, 15, and 25 h after scratch. (g) Relative scratch area quantification (n = 5, mean ± s.d.). Dox combination treatments after siRNA transfection: (h) Short-term cell viability test (n = 4, mean ± s.d.). (i, j) Long-term clonogenic assay (n = 4, mean ± s.d.). One-way ANOVA with Tukey’s post-test. ns, not significant. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.