Investigation of a fusogenic pH-sensitive liposomal system for tumour-targeted intracellular delivery of dinitrobenzamide mustard prodrugs

Abstract

Background: Cancer is the leading cause of death worldwide. While chemotherapy is the mainstay treatment for cancer, non-selective tumour targeting leads to severe side effects. The unique characteristic of solid tumours, such as low pH, hypoxia and the leaky vasculature, may be exploited as triggers to improve therapeutic outcomes. Dinitrobenzamide mustard (DNBM) prodrugs, designed to be selectively activated under tumour hypoxia, were developed as a pharmacological approach for tumour targeting. While they have demonstrated hypoxia selectivity and tumour regression in animals, certain design characteristics, such as poor water solubility, short half-life and poor cellular uptake, limits their potency. In drug delivery, nano-sized PEGylated pH-sensitive liposomes (pPSL) have been investigated as an effective drug carrier (and solubiliser) to confer improved tumour targeting. pPSL containing phosphatidylethanolamine are fusogenic and can achieve preferential cellular uptake, via endocytosis, and endosome escape to avoid lysosomal degradation enhancing its availability to DNA targets relative to DNBM prodrugs alone. Aim: The overall aim of this thesis was to investigate a combination of pharmacological targeting and drug delivery nanotechnology using a novel pPSL delivery system to improve tumour selective targeting and cellular uptake of DNBM prodrugs. The features of solid tumour were exploited. A weakly acidic (SN25860) and basic (SN23816) DNBM prodrug were used as model drugs. Methods: A PEGylated (3 mol%), stable pPSL and non pH-sensitive liposome (nPSL) were developed. The membrane composition of pPSL was based on published reports of 1,2-dioleoyl-sn-glycerol-3-phosphatidylamine (DOPE) and cholesteryl hemisuccinate (CHEMS) (6:4) for pH-responsiveness, with modifications using a high transition temperature (Tc) phospholipid to improve rigidity of the membrane. pH-responsiveness was investigated using a calcein self-quenching assay at pH 7.4, 6.5 and 5.5. Subsequently, SN25860 was loaded using the optimised pPSL. To increase drug loading (DL) and entrapment efficiency (EE), an active loading method (intraliposomal calcium acetate) was used; DL factors were investigated. SN23816 was loaded into the pPSL membrane using a pH gradient. In vitro cytotoxicity of pPSL-SN25860 was investigated using a sulforhodamine B (SRB) assay in a mouse mammary carcinoma cell line (EMT6) transfected with a DNBMactivating nitroreductase gene (nfsB). IC50 values were compared with nPSL-SN25860 and free drug. Cytotoxicity was tested at extracellular pH (pHex) representative of normal tissues (pHex 7.4) and poorly perfused regions of tumours (pHex 6.5). Cytotoxicity of SN23816 was also investigated using an SRB assay, in three Chinese hamster ovarian cell lines, 51D1.3, its homologous recombination repair defective isogenic derivative 51D1 and 51D1 sPOR, which over-expressed a truncated form of the human NADPH:cytochrome P450 oxidoreductase (sPOR), known to be the major SN23816 reductase in hypoxic cells. Free SN23816 was tested in all cell lines to determine hypoxia selectivity; subsequently, pPSL-SN23816 was tested for the effect on cytotoxicity and hypoxia selectivity in 51D1 sPOR. To understand the effect of ionisation on cellular uptake of SN25860 and SN23816, a simulation was modelled based on pKa of the drugs and pH gradient across the cell membrane. A novel quantitative HPLC method was used to measure intracellular and extracellular drug concentrations (Ci and Ce) of EMT6-nfsB cells following exposure to free drug, pPSL and nPSL for 4 h. Flow cytometry and confocal microscopy using calcein (10 mM or 80 mM) loaded liposomes were employed to study cellular uptake, the pathway of endocytosis and endosome escape. Balb/c mice were used to pilot long-circulation of pPSL-/nPSL-SN25860 compared to free drug (n = 2). EMT6-nfsB tumour bearing CD-1 nude mice were used in tissue distribution (n = 3) and antitumour activity, with a clonogenic assay of excised tumour xenografts (n = 7). Results and discussion: An optimised pPSL system consisting of DOPE/CHEMS (6:4) was developed. The addition of a high Tc phospholipid [e.g. 1,2-distearoyl-sn-glycero-3- phosphocholine (DSPC)] and cholesterol improved the rigidity of the membrane in physiological conditions without compromisied pH-responsiveness. The pPSL membrane retained 90% of calcein at pH 7.4, 50% at pH 6.5 and only 20% at pH 5.5; this is promising for targeting pHex in tumours (pH 6.0 - 6.5) and endosome pH (pH 5.0 - 5.5). The optimisation of formulation factors enabled highly drug loaded pPSL (DL = 7.0 ± 0.2%; EE = 76.5 ± 2.4%) for SN25860 and 7.0 ± 0.0 % DL (EE = 75.6 ± 0.3) for SN23816. In vitro cytotoxicity demonstrated that at pH 7.4, the IC50 value of free SN25860 (47.6 ± 1.7 μM) for 18 h drug exposure was similar to that of nPSL-SN25860 (37.6 ± 1.6 μM), however pPSL-SN25860 (2.15 ± 0.33 μM) had a 22- fold decrease in IC50 compared to free drug. For all SN25860 formulations, cytotoxicity was enhanced by ≥ 2-fold at pH 6.5, which indicated that the lower pHex of tumours may be exploited to enhance cytotoxicity of the SN25860 due to the increased proportion of the unionised form. In cytotoxicity investigations of SN23816, free drug demonstrated hypoxia selectivity in all cell lines. SN23816 had the highest increase in potency under anoxic conditions in 51D1 sPOR cells [IC50 (oxic) / IC50 (hypoxic) = 24.73 ± 4.12]. Interestingly, pPSL-SN23816 (DL = 7.0 ± 0.0%) also increased cytotoxicity of the drug under both oxic and anoxic conditions (8.6- and 7.2-fold respectively), without affecting the hypoxia selectivity. HPLC analysis showed that in the weak acid SN25860 had significantly reduced cellular uptake compared to the weak base SN23816 (24-fold decrease in Ci/Ce) at physiological conditions [pHex = 7.4, intracellular pH (pHi) = 7.0], which was consistent with the predicted values. The pPSL formulation increased cellular uptake of both SN25860 and SN23816 to a similar degree, 2.5- and 2.0-folds respectively, which suggested that the mechanism of increasing cellular uptake by pPSL was similar for both drugs. Flow cytometry showed that the pPSL had a 10-fold increase over the nPSL in cellular uptake. For both pPSL and nPSL, 46% and 40% of cellular uptake was attributed to clathrin-mediated endocytosis, respectively. Confocal microscopy confirmed that cellular uptake and endosome escape of pPSL was significantly faster than nPSL. The latter was inhibited by pre-treating cells with omeprazole or chloroquine to increase endosome pH. Overall, cellular uptake studies by HPLC analysis, flow cytometry and confocal microscopy confirmed the hypothesis that pPSL- SN25860 and SN23816 had fusogenic properties due to DOPE in the bilayer, which facilitated endocytosis for efficient cellular uptake as well as achieving endosome escape. In vivo studies of pPSL-SN25860 was selected due to the ability to increase DL, a significant improvement in DL was achieved (31.1 ± 1.3 %) by optimising calcium acetate concentration, increasing drug concentration and maintaining pH of the drug loading medium. A high DL of nPSL-SN25860 (DL = 33.2 ± 1.4%) was also developed and used as a reference. pPSL- and nPSL-SN25860 were able to remain in the systemic circulation, whereas free drug solution was undetectable after 18 h (p < 0.05). This indicated that 3 mol% PEGylation was sufficient for avoiding clearance by the reticuloendothelial system (RES). Furthermore, following intravenous (i.v) injection of the pPSL-SN25860, 0.47 ± 0.08% of the dose was found in tumour after 48 h, while nPSL-SN25860 treatment delivered only 0.06 ± 0.00%; free drug was undetectable. In the liver, no drug was detected for pPSL in contrast to free drug, indicating that pPSL improved safety. Finally, with a single dose, a clonogenic assay of tumour xenografts demonstrated a significant increase in antitumour effect of pPSL-SN25860 compared to drug solution (p = 0.005) and control (saline) (p < 0.0001). Conclusion: The proposed synergistic approach of combining pharmacology and pharmaceutical nanotechnology for tumour targeting successfully improved the cytotoxicity of SN25860 and SN23816 in cancer cells by enhancing intracellular cytoplasmic delivery. Furthermore, the highly drug loaded pPSL was able to improve tumour distribution of SN25860 and have significantly improved antitumour effects. In addition, for the first time a weak acid was successfully loaded into pH-sensitive liposomes with high drug loading.