Pharmacophore hybridisation and nanoscale assembly to discover self-delivering lysosomotropic new-chemical entities for cancer therapy

Abstract

Integration of the unique advantages of the fields of drug discovery and drug delivery is invaluable for the advancement of drug development. Here we propose a self-delivering one-component new-chemical-entity nanomedicine (ONN) strategy to improve cancer therapy through incorporation of the self-assembly principle into drug design. A lysosomotropic detergent (MSDH) and an autophagy inhibitor (Lys05) are hybridised to develop bisaminoquinoline derivatives that can intrinsically form nanoassemblies. The selected BAQ12 and BAQ13 ONNs are highly effective in inducing lysosomal disruption, lysosomal dysfunction and autophagy blockade and exhibit 30-fold higher antiproliferative activity than hydroxychloroquine used in clinical trials. These single-drug nanoparticles demonstrate excellent pharmacokinetic and toxicological profiles and dramatic antitumour efficacy in vivo. In addition, they are able to encapsulate and deliver additional drugs to tumour sites and are thus promising agents for autophagy inhibition-based combination therapy. Given their transdisciplinary advantages, these BAQ ONNs have enormous potential to improve cancer therapy.

Fig. 1. Schematic illustration of the proposed drug design strategy and the current work.

a An interdisciplinary drug design strategy is proposed to integrate the conventional fields of medicinal chemistry and nanomedicine. Drugs are named as one-component new-chemical-entity nanomedicines (ONNs), which are designed according to the strategies of conventional drug design and molecular self-assembly so that they could acquire the advantages from the perspectives of both drug discovery and drug delivery. b The proof-of-concept experiment in this work: discovery of self-delivering lysosomotropic bisaminoquinoline (BAQ) derivatives for cancer therapy. The BAQ derivatives, generated from the hybridisation of lysosomotropic detergents and the BAQ-based autophagy inhibitor, can self-assemble into BAQ ONNs that show enhanced functions in vitro, excellent delivery profiles and significant in vivo therapeutic effects as single agents. Moreover, they also possess high drug-loading efficiency to deliver the additional drug into tumour sites, thus generating a promising application of combination therapy.

Fig. 2. Characterisation of BAQ ONNs.

a Size change of BAQ NPs (10 µM) in acetate buffer with different pH values; data are mean values ± SD; n = 3 independent nanoparticle samples. b The pH-dependent haemolysis induced by BAQ NPs (50 µM, 4 h) in PBS buffer; data are mean values ± SD; n = 3 independent nanoparticle samples. c The pH change of BAQ NPs (1 mM) within hydrochloric acid (HCl, 0.1 M) titration. d, e Representative TEM micrograph at pH 7.4 (d) and pH 5.0 (e); the insets display the size distribution (left) and Tyndall effect (right); experiments were all repeated three times independently. f In vitro drug releasing patterns at pH 7.4 and pH 5.0; data are mean values ± SD; n = 3 independent nanoparticle samples. g The count rate for various concentrations of BAQ NPs in water; The intersection of two lines refers to CACs of BAQ12 NPs (0.45 µg mL⁻¹, 0.76 µM) and BAQ13 NPs (0.15 µg mL⁻¹, 0.25 µM); data are mean values ± SD; n = 3 independent nanoparticle samples.

Fig. 3. BAQ ONNs induced lysosomal disruption and inhibited autophagy in MIA PaCa-2 cells.

a Representative images for cellular uptake of nanoparticles; Dextran-AF488-loaded cells were incubated with DiD-labelled BAQ ONNs for 2 h; experiments were repeated three times independently. b Cells were treated as indicated (10 µM, 2 h) and were stained by LysoTracker Green; experiments were repeated three times independently. c AO staining of cells within the indicated treatments (5 µM, 12 h); experiments were repeated three times independently. d Representative images of Dextran-AF488-loaded cells that were treated as indicated (5 μM, 12 h); experiments were repeated three times independently. e Cathepsin B release from isolated lysosomes after treatments as indicated (25 µM, 12 h); data are presented as mean values ± SD; n = 3. f Western blotting. g Normalised quantification analysis of gel blots in f; data are presented as mean values ± SD; n = 3. h Representative LC3B-GFP images for the indicated 4 h treatments; experiments were repeated three times independently. i Quantification of LC3B-GFP puncta per cell in h; data are presented as mean values ± SD; n = 3. j Representative TEM images of cells that were treated as indicated (2 µM, 48 h); orange rectangle: region of interest; purple arrows: autophagic vesicles; red arrows: lysosomes. k The average diameter of lysosomes; data are presented as mean values ± SD; n = 7. All statistical p values were calculated by one-way ANOVA with the Tukey’s multiple comparison test. ns., not significant; *p < 0.05; **p < 0.01; ****p < 0.0001.

Fig. 4. BAQ ONNs altered the expression of lysosomal genes and caused cell death via apoptosis.

a GSEA demonstrating the enrichment of lysosomal gene sets in MIA PaCa-2 cells treated with BAQ13 NPs (5 µM, 24 h). GSEA was performed with n = 1000 permutations, where p adjust < 0.05 and FDR < 0.05 were considered significant. b Representative upregulated lysosomal genes from a. c Comparison of gene upregulation between Lys05 and BAQ13 NPs. d, e qPCR analysis of V-ATPase genes (d) and Cl⁻ channel genes (e) involved in the indicated treatments (5 μM, 24 h); data are mean values ± SD; n = 3. f Viability curves of cells that were exposed to the 48 h treatments and the corresponding IC₅₀ values; data are mean values ± SD; n = 3. g MIA PaCa-2 (1.5 µM) and HT29 (1.0 µM) cell growth curves within continuous treatments; data are mean values ± SD; n = 3 independent experiments. h Clonogenic assay of MIA PaCa-2 and HT29 cells; n = 3. i Caspase 3/7 activity in MIA PaCa-2 and HT29 cells that were treated for 6 and 12 h, respectively; Data are mean values ± SD; n = 4. j Percentage of apoptotic population of MIA PaCa-2 (left) and HT29 (right) cells that were treated for 24 h. All the statistical p values were calculated by one-way ANOVA with the Tukey’s multiple comparison test; ns not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Fig. 5. The pharmacokinetics, biodistribution and in vivo antitumour effect of BAQ ONNs.

a The plasma concentration-time profiles of DiD-loaded BAQ ONNs and free DiD after intravenous injection; data are mean values ± SD; n = 3. b In vivo and ex vivo biodistribution of BAQ13 NPs in mice bearing HT29 tumours at 24 h post injection. c Quantitative fluorescence intensity of tissues obtained at 12 and 24 h post injection; data are mean values ± SD; n = 3. d MIA PaCa-2 tumour growth curves in mice that were treated as indicated every 3 days; data are mean values ± SD; n = 6 tumours per group. e Body weight of mice during the treatment; data are mean values ± SD; n = 6 mice per group. f Weight of harvested tumours at the end of the treatment; data are mean values ± SD; n = 6 tumours per group. g–j Representative H&E (g), IHC (h), immunoblotting (i) and TEM (j) results of tumours that were harvested at the end of treatments; Blots in i were from three individual tumours of each group; purple arrows in j: autophagic vesicles; experiments in g–j were all repeated three times independently. All statistical p values were calculated by one-way ANOVA with the Tukey’s multiple comparison test; *p < 0.05; ****p < 0.0001.

Fig. 6. BAQ ONNs have dual roles in the combination treatment.

a Establishment of the patient-derived pancreatic cancer stem cell (PCSC) model. b Histological analysis showing the high-level stroma of PCSC tumours; experiments were repeated three times independently. c Viability curves of PCSCs that were treated for 48 h and the IC₅₀ values; n = 3 independent experiments. d AO staining to show the LMP of PCSC that were treated for 12 h; experiments were repeated three times independently. e Immunoblotting analysis of autophagy proteins in PCSC that were treated as indicated (2.5 µM, 48 h); experiments were repeated three times independently. f Synergistic effect of BAQ13 NPs and napabucasin (48 h). g Tumour growth curves in subcutaneous PCSC xenograft model within the indicated treatments every 3 days; data are mean values ± SD; n = 10 tumours per group. h Images of tumours that harvested at end of treatment. i Mice body weight changes during the treatment; data are mean values ± SD; n = 5 mice per group. j Representative images of PCSC tumour sections; experiments were repeated three times independently. k In vivo and ex vivo fluorescence imaging of BAQ13 NPs co-loading with napabucasin and DiD in the PCSC model at 48 h post intravenous injection (10 mg kg⁻¹). l Quantitative fluorescence intensity of tissues in k, data are mean values ± SD; n = 3 mice per group. All statistical p values were calculated by the two-tailed Student’s t test. ns not significant; *p < 0.05; **p < 0.01; ****p < 0.0001.