Peptide-Assisted Intracellular Drug Delivery for Boron Neutron Capture Therapy

Abstract

Background: Glioblastoma is the most aggressive type of brain tumour. The current treatment for glioblastoma is particularly challenging since the tumour is highly infiltrative and has a diffuse growth nature. Boron neutron capture therapy (BNCT) is a promising therapy to selectively damage glioblastoma tissues. The major hurdle for successful BNCT is the requirement for tumour-targeting intracellular delivery of 10B. Sodium borocaptate (BSH) is a clinically approved hydrophilic boron compound with high 10B content but low cellular permeability and tumour-specificity. Nano-sized PEGylated liposomes have been extensively investigated as an effective drug carrier to confer improved tumour targeting. Recently, peptide-functionalised delivery strategies, including conjugated drug molecules or liposomal carriers, have been highlighted for specific intracellular drug delivery. Aim: The overall aim of this thesis was to develop peptide-assisted strategies along with liposomal carriers to specifically deliver sufficient 10B to glioblastoma cells, and therefore enhance the efficacy of BNCT. A cell-penetrating peptide Xentry and a cell targeting peptide cyclic arginine-glycine-aspartic acid-tyrosine-cysteine (c(RGDyC)) were utilised as ligands which were conjugated on the PEGylated liposome surface. The intracellular delivery and tumour-targeting effects were systemically investigated. Methods: Xentry (leucine-cysteine-leucine-arginine-proline-valine-glycine) was first investigated as a carrier for BSH through chemical conjugation. Xentry was synthesised through Fmoc solid-phase peptide synthesis and functionalised with a maleimido group, prior to conjugation with BSH via a Michael Addition reaction. The structures of the products were identified using matrix assisted laser desorption ionization-time of flight mass spectrometer (MALDI-TOF MS) and high resolution electrospray ionisation mass spectrometry (HRESIMS). An isocratic high performance liquid chromatography (HPLC) method was developed for the simultaneous analysis of BSH and the BSH-Xentry conjugate, an ion-pairing reagent was used to confer a desirable retention time for the hydrophilic compound BSH. The composition of mobile phase was optimised using a multiple linear regression model. This HPLC method was validated for rapid analysis of BSH for the formulation development in this project. An optimised PEGylated liposome consisting of DPPG, cholesterol and DSPEPEG2000 at molar ratio of 6.5: 3: 0.5 was developed for BSH delivery. Two different preparation methods, microencapsulation vesicle and thin-film hydration (with freezethaw) methods, were explored to achieve optimal liposomes stability and encapsulation efficiency of BSH. Thereafter, Xentry was utilised as a vector on the surface of the resulting PEGylated liposomes with an aim to enhance the intracellular delivery of BSH. Xentry was conjugated to liposomes via either the side chain of cysteine (XS-LP) or N-terminus (XN-LP). The biocompatibility of the liposomes was assessed by haemolysis assay. Fluorescence intensity quantification and confocal microscopy using calcein (10 mM) loaded liposomes were employed to study the cellular uptake and the internalisation mechanism of XS-LP and XN-LP on a human glioblastoma cell line, U87. To address the limited glioblastoma-specific tissue accumulation as well as the poor cellular penetration of BSH, a novel approach to dual-target glioblastoma vasculature and tumour cells was hypothesised. Expression of integrins, αvβ3 in U87 cells and human umbilical vein endothelial cells (HUVEC), representing tumour angiogenesis, was determined using Western Blotting with a human pancreatic carcinoma cell line (MIA PaCa-2), a human breast cancer cell line (MCF-7) and a mouse macrophage cell line RAW 264.7 as references. PEGylated liposomes were functionalised with a integrin αvβ3 ligand c(RGDyC) peptide (c(RGDyC)-LP) to exploit the overexpression of integrin αvβ3 in both tumour vasculature and tumour cells of glioblastoma. The cellular uptake of c(RGDyC)-LP on those cell models was investigated through fluorescence microscopy and intensity quantification. An in vitro BNCT study was carried out to evaluate the efficacy of BSH containing c(RGDyC)-LP for glioblastoma treatment. Neutron irradiation to the liposome treated cells was conducted in the OPAL reactor in Australian Nuclear Science and Technology Organisation. Cell viabilities were determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay at multiple time point after the irradiation. Results and discussion: The BSH-Xentry conjugate was successfully synthesised with a yield of 54.5% and high purity (94%). An HPLC method for simultaneous analysis of BSH and BSH-Xentry conjugate was established with the use of the ionpairing reagent tetrabutylammonium hydrogensulfate. The retention times of the two analytes were found to be functions of mobile phase composition presented by a multiple linear regression model within the range of aqueous phase volume fraction from 43% to 52% and organic solvent ratios (methanol/acetonitrile) from 1.11 to 9.4. The HPLC method selected for BSH provided a rapid analysis (10 min) and was validated to be highly reproducible and reliable. PEGylated liposomes consisting of DPPG, cholesterol and DSPE-PEG2000 (6.5: 3: 0.5, molar ratio) with desirable size (100-130 nm) was developed for BSH delivery. This formulation was demonstrated to be suitable for BSH encapsulated liposomes as chemical interaction between BSH and the lipid membrane was found to be negligible. Compared with microencapsulation vesicle method, liposomes prepared by thin-film hydration method exhibited a higher encapsulation efficiency of BSH (reached the theoretical maximum, 6%) via passive loading, a slower drug release rate (75% within 8 h versus 91% within 3 h in microencapsulation vesicle method), and minimal drug leakage observed over 2 months (< 5%). Additionally, different freeze-thaw treatments showed similar effects on the liposome properties. The above PEGylated liposomes were successfully conjugated with Xentry in two different ways. XS-LP and XN-LP showed uniform size at 123.6 ± 0.5 nm and 138.9 ± 0.3 nm, respectively. Both liposomes induced negligible haemolysis (below 1.5%) at the lipid concentration up to 0.25 mg/ml, suggesting they had good biocompatibility. Xentry modification on the liposome surface did not significantly increase the cellular uptake as hypothesised. Nevertheless, the results of liposome cellular uptake on U87 cells suggested that peptide conjugation site could significantly affect the cargo internalisation with XN-LP achieving a two-fold increase in cellular uptake compared to XS-LP. Interestingly, a direct translocation, thiol-mediated pathway was found to be involved in the cellular uptake of XN-LP which has free thiols on liposome surface. The Western Blotting results showed that both U87 and HUVEC had stronger expression of integrin αvβ3 than other investigated cell types, MIA PaCa-2, MCF-7 and macrophage RAW 264.7, supporting our hypothesis of simultaneous dualtargeting of both tumour vasculature and tumour cells of glioblastoma through the design of c(RGDyC)-LP. The optimal condition for conjugation of c(RGDyC) to the liposome surface was found to be incubation for 24 h at 22 ˚C; extending the incubation time or changing temperature did not lead to further increase of c(RGDyC) attachment. The degree of cellular uptake of c(RGDyC)-LP correlated with the αvβ3- expression levels of the cells. In contrast, control liposomes without c(RGDyC) showed comparable cellular uptake on different cell types. In the in vitro BNCT study, the c(RGDyC)-LP containing BSH generated more rapid and significant lethal effects to both U87 and HUVEC than the control liposomes or BSH solution. Interestingly, after exposure to neutron irradiation two different types of subsequent cell death, necrosis and apoptosis, was observed in U87 and HUVEC cells, respectively. Conclusion: The cell-penetrating peptide Xentry may not to be the ideal carrier for BSH considering the duration of synthesis, whilst it could be used as a vector by conjugating through its N-terminus to enable liposomal cargo internalisation via direct translocation. Integrin αvβ3 was demonstrated to have a high expression level on both glioblastoma and its vasculature cells, laying a foundation for a new dual-targeting strategy using this ligand c(RGDyC). The c(RGDyC) peptides functionalised liposomes developed in this thesis were demonstrated to have the potential to specifically deliver boron to glioblastoma and its vasculature cells, addressing the major limitation of poor tumour accumulation of 10B which is required in successful BNCT.