The aim of this work was to investigate the relationship between drug-loading and self-emulsification. Several techniques were used to assess the physico-chemical properties of a self-emulsifying formulation formed by GRAS listed excipients with and without a model drug. The micropolarity at the droplet interface was found to be related to the HLB of the surfactant mixtures and it was shown to increase when the drug ibuprofen (6% w/w) was present. IH-NMR showed that this is due to the preferred solubilisation site of the drug; in fact, surfactant-like drugs preferably solubilise at the palisade layer formed by polyoxyethylene units. Phase diagrams showed that the model drug ibuprofen altered the liquid crystal distribution within the SEDDS. Low frequency dielectric spectroscopy confirmed this relationship between drug-loading and phase changes upon dilution. An attempt to correlate in vitro systems with the in vivo scenario was made by measuring the droplet size of the placebo and ibuprofen formulation (6% w/w) using volumetric flask, dissolution apparatus 2 and the novel in vitro Dynamic Gastric Model, DGM (Institute of Food Research, Norwich, UK). The droplet size measured was found to be dependent on media, system used and presence of the drug. Data obtained were compared with the micropolarity of the SEDDS droplets obtained by pyrene fluorescence, and were found to be in good agreement. Furthermore, the results obtained with the DGM showed a clear relationship between digestion and droplet size, with the drug-loaded formulation being more stable than the placebo, which was shown to increase its size as digestion proceeded. The impact of drug loading within a SEDDS was shown to influence the formulation performance significantly. The behaviour of a SEDDS in vivo can not be satisfactorily described by conventional in vitro systems, as it is highly dependent on the motility, biochemistry and physiology of the digestive system.

Fourteen novel comb shaped amphiphilic polymers were successfully synthesised to determine the effect of polymer architecture on the potential of these amphiphilic polymers for hydrophobic drug delivery. Polyallylamine (PAA) was grafted with four different types of hydrophobic pendant groups (Cholesteryl (Ch), (Fmoc), (Dansyl)and (Naphth)). Some amphiphilic polymers were further reacted with methyl orange to form quaternary ammonium moieties. The polymers were characterised by elemental analysis and nuclear magnetic resonance spectroscopy (NMR). In the aqueous environment the amphiphilic polymers formed nano self assemblies with particle size from 99 to 284 nm. The critical aggregation concentration (CAC) of the self assemblies was successfully determined by surface tension measurement. The CAC ranged from the lowest value of 0.093 to the highest 1.5 mgmL-1 (Ch5 and Fmoc5 respectively). The Fmoc and Naphth grafted polymers showed the presence of two CMC values, this phenomenon was due to stacking of the planar hydrophobic ring structures resulting in excimer formation. The theory of excimer formation was confirmed by the observation of peak shifting on the emission spectra of the compounds in water over a large concentration range (0.023 – 3 mgmL-1). The drug loading potential of the polymers was investigated using five model hydrophobic drugs, propofol, prednisolone griseofulvin, etoposide and novel anticancer agent BNIPDaoct. The Ch5 and Dansyl10 showed excellent drug solubilisation capacities. At 6 mgmL-1 the Ch5 achieved propofol solubility 70-fold greater than its aqueous solubility, prednisolone, griseofulvin and etoposide solubility’s were increased 20-fold, 30-fold and 7-fold respectively. Similarly at 6 mgmL-1 the Dansyl10 achieved a 200-fold increase on the aqueous solubility of propofol and increased the solubility of prednisolone, griseofulvin and etoposide by 100-fold and 400-fold and 12-fold respectively. The Ch5 (at 1 mgmL-1) was also used to solubilise the novel anticancer agent Bisnaphthalimidopropyl diaminooctane (BNIPDaoct) which was otherwise insoluble achieving a solubilisation of 0.3 mgmL-1. The sizes of the optimal formulations differed greatly for both modified polymers. This was possibly due to the varying architectures of both the drug and the modified polymers and their ability to expand the hydrophobic core and shield the drugs from the ‘hostile’ aqueous environment. The in vitro drug release profiles, showed controlled release of the hydrophobic drugs from the core of the nano aggregates (Ch5 and Dansyl10), the time span for 100% of the drugs to be released ranged from 48- 96 h. Biological characterisation of the polymers found that most of the polymers showed negligible haemolytic activity over the concentrations tested (0.05 – 1 mgmL-1), the IC50 values for the cytotoxicity assay ranged from 0.01740 - 0.05585 mgmL-1 on Caco-2 cells (Fmoc5 to QCh5 respectively). The quaternized polymers showed a slightly better safety profile than their unquaternized counterparts, despite exhibiting low drug solubilisation capacities. The optimal formulations of Ch5 and Dansyl10 loaded with etoposide and BNIPDaoct were tested for their cytotoxicity in vitro on Caco-2 and HEK293 cells. All formulations were capable of lowering the IC50 values when compared with the free anticancer drugs, thus increasing their therapeutic effect. The Ch5 decreased the solubility of etoposide 2.2-fold and BNIPDaoct 1.3-fold on Caco-2 cells, with Dansyl10 achieving a 14 -fold and 16–fold reduction respectively. In vivo oral administration of Ch5 and Dansyl10, griseofulvin formulations demonstrated significantly enhanced the absorption of griseofulvin absorption in rats compared with griseofulvin in water (8.89-fold and 5.20-fold increase respectively on total concentration of griseofulvin solubilised over 24 h study). A formulation of Ch5, BNIPDaoct was also shown to significantly decrease the tumour growth when treated on tumour bearing nude mice over a 4 week period. This is the first time these novel PAA’s grafted with cholesteryl and dansyl have shown promising potential in hydrophobic drug delivery.

The effect of crystallisation solvent on crystal morphology and mechanical properties of ibuprofen was studied. Intermolecular potentials were calculated to assess the relative growth rates of crystal faces of ibuprofen. Comparison of the predicted morphology with solution-"grown crystals clarified the interaction of solvent with the growing crystal. Single crystal dissolution and x--ray techniques indicated that crystals grown from non--polar solvents were more strained. In conjunction with solution NMR results (which showed differences in molecular association in solvents of varying polarity) this led to two different growth mechanisms being proposed. The crystal structure of resolved S(+) ibuprofen confirmed that enantiomeric dimers present in non-polar solvents could act as additives, selectively decreasing the growth rate at specific crystal faces, and introducing lattice strain. Deformation processes of ibuprofen did not appear to be changed by the solvent of crystallisation. However, crystals grown from non-polar solvents fractured readily. In conjunction with the unfavourable acicular morphology this is likely to be a source of processing problems of industrially produced crystals.

The efflux of drugs, particularly multi-drug resistant (MDR) drugs, by transporters such as P-glycoprotein provides an additional hurdle to achieving high levels of drug bioavailability. The aim of this project was to investigate drug-binding peptides as potential delivery tools, which might circumvent drug efflux by transporters such as Pglycoprotein. Popkov et 01 1998 reported a number of phage-displayed peptides (I Omers), which bound to doxorubicin. A literature search was carried out and several other potential drug-binding sequences were identified. Peptides were synthesised and then screened for drug-binding and their effect on drug interactions with P-glycoprotein. Due to the small and hydrophobic nature of the drug and peptide molecules, both homogeneous and heterogeneous drug-peptide binding assays were investigated. Drugpeptide binding was identified in one sequence using circular dichroism, size exclusion HPLC, equilibrium dialysis and a new biosensor, the Farfield AnalightBi0200. Competition assays using the fluorophores rhodamine 123 and calcein-AM suggested that MDR drugs formed complexes with these fluorophores. Drug-peptide aggregates were also identified by particle sizing, using photon correlation spectroscopy. Drug-P-glycoprotein interactions were investigated using a cell-based competition assays, cytotoxicity and MDCK cell monolayer transport studies. All the potential drug-binding peptide sequences were screened using a calcein-AM competition assay in PGP over-expressing cells (CHO CHRC5), whereby calcein-AM and drug (± peptide) compete to interact with P-glycoprotein. Only the Popkov-reported sequences were found to reduce drug-P-glycoprotein interactions. However, these sequences did not enhance the cytotoxicity of doxorubicin or vinblastine (cytotoxic P-glycoprotein substrates) in Caco-2 or CHO CHRC5 cells. MDCK monolayer studies found the peptide to have a slight but not significant effect on 3H vinblastine basal to apical transport. This difference may be due to greater sensitivity of the calcein AM compared to the monolayer transport assay. With a view to enhancing the delivery properties of the drug-binding peptides, new sequences containing cell-penetrating motifs were synthesised. The Penetratin sequence was synthesised in tandem to a drug-binding peptide sequence, whilst the Simian virus 40 motif was added to the peptide using a linker. Both these peptides gave similar results in the calcein-AM competition assay as the drug-binding peptide alone. The ability of the peptides to penetrate cell membranes was tested using a calcein-liposome leakage assay. Both the new sequences were slightly lytic, but complete liposome lysis particularly resulted when tamoxifen was added. Interestingly, a similar result was achieved with the drug-binding peptide sequence alone. Some drug-binding peptides appear to affect the behaviour of P-glycoprotein in suspended-cell transport models, when assayed by calcein-AM. However, the interaction between drugs and peptides in other cell-based models appears more complex.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.