Hybrid Nanoparticle Engineered with Transforming Growth Factor -β1-Overexpressed Extracellular Vesicle and Cartilage-Targeted Anti-Inflammatory Liposome for Osteoarthritis

Abstract

Extracellular vesicles (EVs) possess the characteristics of their parent cells, based on which various studies have actively investigated treatments for diseases using mesenchymal stem cell-derived EVs due to their regenerative activity. Furthermore, in recent years, there have been significant efforts to engineer EVs to improve their native activities and integrate additional functions. Although both endogenous and exogenous methods are used for engineering EVs, endogenous methods may pose the problem of administering substances to cells undergoing metabolic changes, which can cause potential side effects. In addition, exogenous methods may have the limitation of losing beneficial factors inside EVs due to membrane disruption during engineering processes. Surface modification of EVs may also impair efficiency due to the presence of proteins on the EV surface. Therefore, in this study, a stable and efficient engineering method was achieved through the ethanol-mediated hybridization of EVs and functionalized lipid nanoparticles (LNPs) with a fusogenic lipid component. During hybridization, the internal bioactive factors and targeting moiety were maintained to possess the characteristics of both LNPs and EVs. The Ab-Hybrid, which was successfully synthesized through hybridization with nicotinamide-encapsulated and Col2A1 antibody-modified liposome and Transforming growth factor-β1 (TGF-β1)-overexpressed EVs, was administered to osteoarthritis (OA)-induced rats undergoing the destabilization of the medial meniscus surgery. Ultimately, the Ab-Hybrid demonstrated excellent chondroprotective and anti-inflammatory effects with targeting and long-lasting properties in OA lesions. We anticipate that this approach for manufacturing hybrid particles will serve as a valuable EV engineering method and a versatile platform technology applicable to various diseases.

Keywords: extracellular vesicles; hybrid nanoparticle; liposome; mesenchymal stem cell; osteoarthritis.

Scheme 1. Schematic Diagram

(A) Fabrication of hybrid nanoparticles and their application in osteoarthritis therapy. (B) Increased exposure of Col2 during the progression of osteoarthritis, along with the enhanced targeting efficiency and retention time of hybrid nanoparticles.

Figure 1

Confirmation of TGF-β1 overexpression in TMSC (TGF-β1 TMSC) and TGF-β1 TMSC-derived EVs. (A) Schematic representation of CRISPR/Cas9-mediated knock-in of TGF-β1 into a safe-harbor site (AAVS1) on the TMSC chromosome. Evaluation of transgenic expression of TGF-β1 with (B) immunocytochemistry and (C) Western blot analysis. (D) Size distribution of EV derived from TMSC and TGF-β1 TMSC. (E) Western blot analysis of EV derived from TMSC and TGF-β1 TMSC with EV positive markers, CD63 and TSG101, and a negative marker ApoA1. (F) Morphologies of EV imaged with TEM. Scale bars equal to 100 nm. (G) Level of TGF-β1 in cell culture media and EVs detected by ELISA. (H) Representative images of wound healing assays to evaluate wound healing effects of EVs and calculated wound healing rates of EVs (values are presented as mean ± SD (n = 3), and statistical significance was obtained with one-way analysis of ANOVA with Tukey’s multiple comparison post-test (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001)).

Figure 2

Nicotinamide (NAM) effect on primary rat chondrocytes. (A) Cell viability in various concentration of NAM. (B) Gene expression levels of Col2A1, MMP13, NF-kB, and IL-6. C. Illustration of the fabrication process of NAM-loaded liposome nanoparticles (N-LNP; LNP). (D) Size distribution of LNP. (E) Loaded amount of NAM in LNP quantified with HPLC. (F) Morphologies of LNP imaged with TEM. Scale bars equal to 100 nm (values are presented as mean ± SD (n = 3), and statistical significance was obtained with one-way analysis of ANOVA with Tukey’s multiple comparison post-test (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001)).

Figure 3

Ethanol-mediated hybridization between EVs and LNPs. (A) Illustration of the FRET-based lipid mixing assay. (B) FRET intensity in various EV and LNP particle ratios (PRs). (C) Changes in fluorescence intensity at each wavelength during EV-LNP hybridization. (D) Zeta potential of EV, LNP, and N-Hybrid. (E) Loaded amount of NAM and TGF-β1 in N-Hybrid quantified with HPLC and ELISA, respectively. (F) Size distribution of N-Hybrid. (G) Morphologies of N-Hybrid imaged with TEM. Scale bars equal to 100 nm. (H) Western blot analysis of N-Hybrid with EV positive markers, CD63 and TSG101, and a negative marker, ApoA1 (values are presented as mean ± SD (n = 3), and statistical significance was obtained with one-way analysis of ANOVA with Tukey’s multiple comparison post-test (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001)).

Figure 4

Functionality of hybrid nanoparticle. (A) Representative images of wound healing assays to evaluate wound healing effects of EV, LNP, and N-Hybrid, and quantification data for wound healing rates of EV, LNP, and N-Hybrid. (B) Gene expression levels of Col2A1, MMP13, NF-kB, and IL-6. (C) Fluorescence-based immunocytochemistry for MMP13 expression. (D) Fluorescence-based immunocytochemistry for NF-κB expression (values are presented as mean ± SD (n = 3), and statistical significance was obtained with one-way analysis of ANOVA with Tukey’s multiple comparison post-test (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001)).

Figure 5

Antibody engineering for functionalization of hybrid. (A) Illustration of antibody engineering process and Ab-Hybrid. (B) Change in fluorescence intensity at each wavelength during hybridization of EV and Ab-LNP. (C) Size distribution of Ab-LNP and Ab-Hybrid. (D) Morphologies of Ab-LNP and Ab-Hybrid imaged with TEM. Scale bars equal to 100 nm. (E) Western blot analysis of Ab-Hybrid with EV positive markers, CD63 and TSG101, and a negative marker, ApoA1. (F) Zeta potential of EV, Ab-LNP, and Ab-Hybrid. (G) Represented images visualizing the internalization of particles at primary rat chondrocytes (PRCs). Scale bars equal to 50 μm. (H) Quantitative value of the fluorescence intensity of particles attached to the PRC (values are presented as mean ± SD (n = 3), and statistical significance was obtained with one-way analysis of ANOVA with Tukey’s multiple comparison post-test (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001)).

Figure 6

Ex vivo and in vivo particle tracking study with N-Hybrid and Ab-Hybrid. (A) Image of particle affinity with damaged femoral condyle imaged with fluorescence-labeled organism-bioimaging instrument (FOBI). (B) Quantitative graph of fluorescence intensity. (C) Illustrated image of DMM surgery and the time schedules of in vivo particle tracking analysis. (D) Time-dependent knee images of rats injected with DiD-labeled N-Hybrid and Ab-Hybrid 6 weeks after DMM modeling (values are presented as mean ± SD (n = 3), and statistical significance was obtained with one-way analysis of ANOVA with Tukey’s multiple comparison post-test (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001)).

Figure 7

Histological image and gene/protein expression levels of the in vivo analysis. (A) Time schedules of in vivo analysis. (B) Representative images of the hematoxylin and eosin (H&E) and safranin O-fast green (S/O) staining of the samples collected at 8 weeks of postsurgery. (C) Quantitative graph of S/O staining intensity and OARSI score based with histological image. (D) Gene expression levels of the in vivo samples of 8 weeks postsurgery. (E) Western blot analysis for the in vivo samples 8 weeks after surgery. (F) Quantitative intensity graph of the Col2A1 and MMP13 normalized with β-actin intensity (values are presented as mean ± SD (n = 3), and statistical significance was obtained with one-way analysis of ANOVA with Tukey’s multiple comparison post-test (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001)).