Biomimetic lipidic self-assembly materials for encapsulation and controlled release of protein therapeutics

Biopharmaceuticals, including therapeutic proteins and peptides, represent a rapidly growing sector of the pharmaceutical industry due to their adaptability and ability to treat many diseases. Protein-based drugs can be highly specific, with reduced adverse side-effects, and maybe less like to induce an immune response.  Protein-based vaccines and diagnostics are also becoming increasingly common. However, such drugs are associated with numerous issues including high cost and low stability, particularly in vivo.  Oral delivery of protein and peptide-based drugs has so far been unsuccessful although many attempts have been made. Lipid-based nanomaterials of cubic symmetry (lipidic bicontinuous cubic phases) have demonstrated a range of advantages as drug delivery vehicles.  These materials can be used both in bulk form (typically topical delivery) or dispersed into sub-micron particles (cubosomes) for subcutaneous injection delivery. They have a large internal surface area, and their amphiphilic nanostructure allows them to encapsulate hydrophobic and hydrophilic drugs of a wide range of molecular weights.  They have specific advantages for protein based drugs as their fundamental lipid bilayer structure mimics the lipid bilayer environment of the cell membrane and may assist in the retention of the protein secondary structure and activity, particularly for membrane-bound proteins.  Research presented in this thesis aims to investigate the potential advances for biopharmaceutical drug delivery in vitro and in vivo using the lipidic bicontinuous cubic phase. This thesis combines a solid physicochemical characterisation of the drug encapsulated lipid materials and nanoparticles, with in vitro and in vivo studies to determine the interaction of the nanoparticles with cells, and their use as drug delivery vehicles for protein drugs in vivo.  Encapsulation of three different peptides: oxytocin and somatostatin (Chapter 2) and insulin (Chapter 4) on the nanostructure of the lipidic cubic phase was determined using SAXS.  The cubic phase architecture was retained up to reasonably high protein loading: 40 mg/ml for oxytocin, 30 mg/ml for somatostatin and 10 mg/ml for insulin.  For oxytocin, which is small and hydrophilic, and therefore located in the water channel region, no changes to the lipidic cubic phase matrix were observed. For somatostatin, which is more associated with the lipid membrane, the addition of peptide both increased the lattice parameter and resulted in a broadening of Bragg diffraction peaks, suggesting an increase in disorder within the sample.  For insulin, the effect on the cubic architecture was related to the geometry of the water channels relative to the size of the protein.  Encapsulation in monoolein (MO), where the insulin dimer is anticipated to fit easily within the water channel resulted in no changes to the cubic matrix, whereas encapsulation in the smaller water channels of the phytantriol (PT) cubic phase resulted in an increase in lattice parameter, presumably due to the cubic phase swelling to accommodate the insulin dimer. In all cases, encapsulation had minimal effect on the secondary structure of the protein, as determined using circular dichroism (CD) analysis.   Release of somatostatin and oxytocin from the lipidic cubic phase was shown to be diffusion-controlled over the early stages (~ 8 hours). Over longer time periods, peptide monomers contained within the cubic phase were demonstrated to self-assemble into fibrils.  This was associated with a large reduction in release rates.  This opens up the possibility of a two-phase release profile for these neuropeptides from the cubic phase with initial fast diffusion-controlled release followed by sustained release of the monomer over a period of several days.  Diffusion of insulin from the cubic phase was also shown to be diffusion controlled.  Diffusion coefficients determined from release data were compared with those generated from fluorescence recovery after photobleaching (FRAP) analysis with good agreement found. Results showed insulin encapsulated in PT had a faster release rate when compared to MO. MO encapsulated insulin was found to have a slower but more consistent release over the same period with more than 30% of encapsulated insulin released over 48 hours. The stability of encapsulated proteins and peptides against enzymatic degradation was tested to determine their potential efficacy for oral or injection delivery. CD analysis was used to track the secondary structure of encapsulated insulin in the presence of a gut enzyme (chymotrypsin), either in solution or encapsulated within a cubic phase.  In solution, following the addition of chymotrypsin the secondary structure of insulin changes quickly over a 12 min period consistent with destruction of the alpha-helices.  In contrast, the same protein, when encapsulated in lipidicic cubic phase (LCP), showed markedly reduced changes to the secondary structure over two hours when exposed to chymotrypsin. This demonstrates the ability of the cubic phase to protect encapsulated proteins against enzymatic degradation over a physiologically relevant timescale.  In order to utilise these materials in in vivo trials, cellular toxicity of cubosomes based on the monoacylglycerol lipids MO and monopalmitolein  (MP), as well as the branched-chain lipid PT, was tested using both confocal microscopy and fluorescence activated cell sortinging (FACS) analysis against ouabain-resistant (STO) and macrophage cell lines.   The toxicity of the monoacylglycerol lipids was relatively low up until 200 µg/ml.  The toxicity of the PT cubosomes was much higher with 100% cell death at 50 µg/ml. The cellular uptake of cubosomes into STO fibroblast and macrophage cell lines was investigated using live-cell imaging on a confocal microscope. The uptake mechanism was shown to depend on the particular cell-line used.  Results for uptake into fibroblast cells were consistent with a fusion mechanism of interaction. In contrast, repeating fluctuations in the fluorescence profile observed during uptake of cubosomes into macrophage cells are consistent with endocytosis of the cubosome, followed by transfer to the lysosomes and rapid recycling.  Finally, oral delivery of proteins contained within the lipidic cubic phase was assessed in vivo using rats.  As the lipidic cubic phase is rapidly degraded within the stomach, it was contained within a capsule having an enteric coating which is designed to pass intact through the stomach and break down in the small intestine.  A preliminary trial was run using green fluorescent protein (GFP) with promising results - an efficacy of 62 % relative to subcutaneous (SC) injection delivery was determined based on blood plasma fluorescence readings.  Based on these promising results, subsequent animal trials investigated the oral delivery of insulin to diabetic Sprague Dawley rats using the same formulation system.   Both fast-acting and slow-acting forms of insulin were tested with good efficacy determined for both. The hydrophobic lipid-conjugated slow-acting insulin (Levemir) contained within an enteric capsule with a thin coating was found to have the highest efficacy of all systems tested with an efficacy (relative to SC injection) of more than 150%. Results presented in this thesis confirm that lipidic bicontinuous cubic phases are highly prospective drug delivery vehicles for protein and peptide-based drugs.  Good encapsulation efficiencies were measured, with minimal change to the lipid nanostructure for most proteins.  The secondary structure of the protein, and hence the activity, was generally unaffected by encapsulation.  Release from these materials was generally diffusion-controlled at least over earlier time periods.  Protection of the encapsulated protein from enzymatic attack was demonstrated.  Finally, the combination of a lipidic cubic phase and an enteric capsule was shown to successfully achieve oral delivery of insulin, with very high efficacy particularly for lipid-conjugated slow-acting insulin.