Inhalation of acidic nanoparticles prevents doxorubicin cardiotoxicity through improvement of lysosomal function

Abstract

Doxorubicin (Dox) is an effective anticancer molecule, but its clinical efficacy is limited by strong cardiotoxic side effects. Lysosomal dysfunction has recently been proposed as a new mechanism of Dox-induced cardiomyopathy. However, to date, there is a paucity of therapeutic approaches capable of restoring lysosomal acidification and function in the heart. Methods: We designed novel poly(lactic-co-glycolic acid) (PLGA)-grafted silica nanoparticles (NPs) and investigated their therapeutic potential in the primary prevention of Dox cardiotoxicity in cardiomyocytes and mice. Results: We showed that NPs-PLGA internalized rapidly in cardiomyocytes and accumulated inside the lysosomes. Mechanistically, NPs-PLGA restored lysosomal acidification in the presence of doxorubicin or bafilomycin A1, thereby improving lysosomal function and autophagic flux. Importantly, NPs-PLGA mitigated Dox-related mitochondrial dysfunction and oxidative stress, two main mechanisms of cardiotoxicity. In vivo, inhalation of NPs-PLGA led to effective and rapid targeting of the myocardium, which prevented Dox-induced adverse remodeling and cardiac dysfunction in mice. Conclusion: Our findings demonstrate a pivotal role for lysosomal dysfunction in Dox-induced cardiomyopathy and highlight for the first time that pulmonary-driven NPs-PLGA administration is a promising strategy against anthracycline cardiotoxicity.

Keywords: autophagy; cardiotoxicity; doxorubicin; lysosomes; nanoparticles.

Figure 1

PLGA-grafted silica particles characterization. (A) Scheme showing the steps of PLGA grafting on silica particles (B) Transmission electron microscopy (TEM) (Scale Bar = 50 nm) and (C-D) Size distribution histogram of submicronic PLGA-grafted silica particles (SMPs-PLGA) with (C) length and (D) width. (E) Transmission electron microscopy (TEM) (Scale Bar = 50 nm) and (F) Size distribution histogram of PLGA-grafted silica nanoparticles (NPs-PLGA) (G) Fourier-transform infrared spectroscopy (FTIR) of NPs, SMPs-PLGA and NPs-PLGA. The x-axis represents the wavenumber (cm^-1) and the y-axis represents the percentage of transmittance. The spectra of naked particles (NPs) show the presence of well-known Si-O-Si (1070 and 795 cm^-1), Si-O (461 cm^-1), and Si-O-H (3422 cm^-1) bands and a strong peak at 1630 cm^-1 attributed to water molecules. Additional bands observed on SMPs-PLGA and NPs-PLGA correspond to ν_C-H (2945 cm^-1) and δ_C-H (1550, 1473, 1430 cm^-1) of both APTES and PLGA and to ν_C=O of PLGA (1750 cm^-1).

Figure 2

PLGA-grafted nanoparticles are driven to lysosomal compartment in NRVMs. (A) Intracellular uptake rate of NPs-PLGA-FITC in NRVMs (n = 3). (B) Left panel: Representative confocal images of NRVMs transfected with LAMP1-RFP (red) and incubated with NPs-PLGA-FITC (green) for 8 and 24 h. Hoechst 33258 (blue) was used as counterstaining. Scale Bar = 10 μm. Right panel: Intensity profile graphs of NPs-PLGA-FITC colocalization with LAMP1-RFP. (C) 3D reconstruction of stack images of NRVMs transfected with LAMP1-RFP (red) and incubated with NPs-PLGA-FITC (green) for 24 h. Hoechst 33258 (blue) was used as counterstaining. Scale Bar = 5 μm for whole images, 1 µm for magnified areas. (D) Left panel: Orthogonal projection of stack images of NRVMs transfected with LAMP1-RFP (red) and incubated with NPs-PLGA-FITC (green) for 24 h. Transmitted light (TL) was used to determine cell boundaries on the whole image and Hoechst 33258 (blue) was used as counterstaining. Scale Bar = 5 μm for whole image, 2 µm for magnified areas. Right panel: Manders' overlap coefficient of green fluorescence overlapping red one and red fluorescence overlapping green one (n = 6). (E) Lysosomal immunoisolation in NRVMs transfected with TMEM192-3xHA plasmid. Left panel: Representative immunoblots for protein markers of various subcellular compartments in whole cell lysates (input) and purified lysosomes (lysosomal fraction) from NRVMs incubated or not with NPs-PLGA-FITC. Right panel: quantification of green fluorescence in pure lysosomes from NRVMs incubated or not with NPs-PLGA-FITC (n = 3). Data are expressed as means ± SEM (**p < 0.01, ***p < 0.001 vs t = 0 or No NPs).

Figure 3

NPs-PLGA prevent Dox-induced lysosomal dysfunction. (A-C) H9C2 were pre-incubated O/N with Veh or 25 µg/mL of NPs or NPs-PLGA and treated with doxorubicin (1 µM, 10 h). (A) Left panel: Representative images of Acridine Orange red fluorescence on H9C2. Hoechst 33258 (blue) was used as counterstaining. Scale Bar = 10 µm. Right panel: Fluorimeter measurements of Acridine Orange red staining (n = 4). (B) Fluorimeter measurements of Lysosensor yellow/blue staining on H9C2 (n = 5). Excitation was measured at 360 nm and the ratio of emission 440/540 nm was calculated. (C) Fluorimeter measurements of Cathepsin D activity fluorescence on H9C2 (n = 6) (D-E) NRVMs were pre-incubated O/N with Veh or 200 µg/mL of NPs, SMPs, NPs-PLGA or SMPs-PLGA and treated with doxorubicin (1 µM, 10 h). (D) Fluorimeter measurements of Lysosensor yellow/blue staining on NRVMs (n = 6). (E) Fluorimeter measurements of Acridine Orange red staining on NRVMs (n = 4). Data are expressed as means ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 vs Veh; #p < 0.05, ##p < 0.01 vs Dox).

Figure 4

Intratracheal administration of NPs-PLGA leads to cardiac lysosomes addressing in mice. (A) Schematic representation of NPs-PLGA administration via intratracheal nebulization in mice. (B-C) Fluorescence of NPs-PLGA-FITC in (B) lungs and (C) hearts of mice treated with Veh or 1 mg/kg of NPs-PLGA-FITC through intratracheal nebulization for 10', 30', 1 h or 4 h. Data are normalized per mg of tissue. (n = 3) (D) 3D reconstruction of stack images of heart tissues from mice collected 1 h after intratracheal nebulization of Veh or 1 mg/kg of NPs-PLGA-FITC. Green: NPs-PLGA-FITC, red: alpha-actinin, blue: DAPI. Scale bar = 5 µm. (E) Transmission Electron Microscopy of cardiac lysosomes from mice after 1h of Veh or NPs-PLGA-FITC (1 mg/kg) administration via intratracheal nebulization. Data are expressed as means ± SEM (*p < 0.05, **p < 0.01 vs Veh).

Figure 5

NPs-PLGA avoid autophagy alterations after Dox treatment. (A) Schematic representation of the experimental design in mice treated with Veh or Dox (5 mg/kg) ± NPs-PLGA (0.5 or 1 mg/kg) for 24 h. Dox and NPs-PLGA were administered at the same time. (B) Upper panel: Representative confocal images of heart tissues from mice 24 h after Veh or Dox (5 mg/kg) ± NPs-PLGA (0.5 or 1 mg/kg) administration. Green: alpha-actinin, red: LC3, blue: DAPI. Scale bar = 10 µm. Lower panel: Quantification of the number of LC3-positive dots per 0.02 mm² (n = 3). (C) Upper panel: Representative immunoblots and lower panel: quantifications of p62 and LC3II expression in cardiac homogenates of mice 24 h after Veh or Dox (5 mg/kg) ± NPs-PLGA (0.5 or 1 mg/kg) administration (n = 3). GAPDH was used as loading control. (D) Upper panel: Double immunofluorescence imaging of RFP-GFP-LC3-transfected NRVMs pre-incubated O/N with Veh or 200 µg/mL of NPs or NPs-PLGA and treated with doxorubicin (1 µM, 16 h). Scale bar = 10 µm. Lower panel: Quantification of yellow puncta (AP, autophagosomes) and red puncta (AL, autolysosomes) for each condition (n = 6). Data are expressed as means ± SEM (*p < 0.05, **p < 0.01 vs Veh; #p < 0.05, ##p < 0.01 vs Dox).

Figure 6

NPs-PLGA limit Dox-induced mitochondrial dysfunction. (A) Experimental design in mice treated 4 times with once-per-week injections of Veh or Dox (5 mg/kg) ± NPs-PLGA (1 mg/kg) at the same time, and analysed at 28 days. (B-C) Immunoblots from cardiac homogenates of mice at 28 d with: (B) OXPHOS complexes (Left panel: Representative immunoblots and right panel: quantifications) (n = 8-12). GAPDH was used as a loading control. (C) 4-HNE (Left panel: Representative immunoblots and right panel: quantifications) (n = 8-12). GAPDH was used as a loading control. (D-F) H9C2 were pre-incubated O/N with Veh or 25 µg/mL of NPs or NPs-PLGA, and treated with Veh or Dox (1 µM). (D) Quantification of mitoROS assessed by mitoSOX fluorescence on H9C2 treated with Veh or Dox (1 µM, 10 h) (n = 5) (E) Quantification of mitochondrial depolarization assessed by fluorescence of JC-1 red aggregates/green monomers ratio on H9C2 treated with Veh or Dox (1 µM, 10 h) (n = 5) (F) Left panel: Representative immunoblots and right panel: quantifications of p-p53 and cleaved caspase 3 protein expression on H9C2 treated with Veh or Dox (1 µM, 24 h). GAPDH was used as a loading control. (n=6) (G) Quantification of caspase 3 activity in NRVMs treated with Veh or Dox (1 µM, 24 h) (n = 4). Data are expressed as means ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 vs Veh; #p < 0.05, ##p < 0.01, ###p < 0.001 vs Dox).

Figure 7

Dox-induced cardiac adverse remodeling and dysfunction are prevented upon NPs-PLGA treatment. (A-F) Evaluation of cardiac parameters after 4 once-per-week injections of Veh or Dox (5 mg/kg) ± NPs-PLGA (1 mg/kg). Dox and NPs-PLGA were administered at the same time. (A) Cardiac remodeling was evaluated by determining the Heart weight-to-Tibia length ratio (n = 8-12). (B) Left panel: Representative images of Vinculin staining (Scale bar = 25 µm) and right panel: quantification of cardiomyocyte area in heart sections (n = 8-12). (C) Left panel: Representative images of TUNEL staining (Scale bar = 50 µm) and right panel: quantification of TUNEL positive nuclei in heart sections (n = 8-12). (D) Left panel: Representative images of Sirius Red staining (Scale bar = 25 µm) and right panel: quantification of collagen content in heart sections (n = 8-12). (E-F) Cardiac function was assessed by echocardiographic parameters with (E) Fractional Shortening (n = 8-12) and (F) Ejection Fraction (n = 8-12). Data are expressed as means ± SEM (*p < 0.05, ***p < 0.001 vs Veh; #p < 0.05, ##p < 0.01, ###p < 0.001 vs Dox).

Figure 8

NPs-PLGA do not alter Dox anti-cancer effect. (A-C) Proliferation curves of cancer cells incubated with Veh or NPs-PLGA (200 µg/mL) and treated with Veh or Dox (5 µM) with (A) rat urothelial carcinoma AY27 cells (B) mouse mammary carcinoma 4T1 cells and (C) human histiocytic lymphoma U937 cells (n = 4). (D-F) Apoptosis measured by annexin fluorescence normalized to cell number in cancer cells incubated with Veh or NPs-PLGA (200 µg/mL) and treated with Veh or Dox (5 µM) with (D) rat urothelial carcinoma AY27 cells (E) mouse mammary carcinoma 4T1 cells and (F) human histiocytic lymphoma U937 cells (n = 4). Data are expressed as means ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 for Dox vs Veh, #p < 0.05, ##p < 0.01, ###p < 0.001 for NPs-PLGA+Dox vs Veh, $p < 0.05 for NPs-PLGA vs Veh).

Figure 9

Role of lysosomal dysfunction and benefits of NPs-PLGA in Dox cardiotoxicity. Dox induces an increase in lysosomal pH, contributing to organelle dysfunction, autophagic flux blockade and exacerbating mitochondrial alterations, all of these participating in cardiotoxicity. Co-administration of NPs-PLGA with Dox prevents lysosomal alkalinization through restoration of lysosomal acidity, thereby restoring autophagic flux and preserving mitochondria function, leading to cardioprotection. This figure was created with BioRender.com.