Abstract
Purpose: The increasing interest in new generation vaccines is based upon utilising highly purified proteins and peptides as antigens. However, a disadvantage of these subunit vaccines is that they are often poorly immunogenic. Therefore there is a need to develop new formulation strategies which can generate the desired immune responses and are both safe and efficacious. One important formulation strategy involves incorporating antigens in polymeric nanoparticles, while another formulation strategy involves the dispersion of antigens in a sustained release carrier with the aim of increasing the size of the immune response generated and perhaps avoiding the need for multiple immunisations. In this thesis the aim was to combine these two approaches by dispersion of polymeric nanoparticles loaded with the antigen into a sustained release delivery systems (a self gelling microemulsion). Two polymeric nanoparticles (poly(ethylcyanoacrylate), PECA and chitosan nanoparticles, CNP) were dispersed in self gelling microemulsion (ME) templates. Ovalbumin (Ova) and Quil A were used as model antigen and adjuvant respectively. Biocompatible microemulsions were developed by establishing pseudoternary phase diagrams and these were then characterised at 37 ˚C. Lamellar liquid crystalline gel regions were found adjacent to the microemulsion regions in the phase diagrams. Upon addition of water or body fluids a phase transition from microemulsions to liquid crystalline systems may occur, which may make these systems suitable for sustained antigen delivery, as the rate of diffusion within the liquid crystalline phase is slower as compared to liquid vehicles. The study investigated the influence of the oil and water composition of the microemulsions on viscosity and release of antigen dispersed in the microemulsion in molecular form and incorporated into nanoparticles from the microemulsions and liquid crystalline gels. The ability of these formulations to generate immune responses towards Ova was also investigated in vivo.
Methods: The ME components consisted of isopropyl myristate, lecithin, ethanol, water and either decyl glucoside (DG) or capryl-caprylyl glucoside (CCG). The selected microemulsions had a surfactant: water (S:W) ratio of 9:1 and a surfactant: oil (S:O) ratio of 5.2: 4.8. Formulations were loaded with fluorescently labelled Ova (FITC-Ova) and used as polymerisation templates for the preparation of PECA nanoparticles by interfacial polymerisation. CNP were prepared separately by a precipitation/coacervation method facilitated by sodium sulphate. The viscosity of one phase and two phase liquid crystalline gels (with and without nanoparticles incorporated) was determined using a cone and plate rheometer. A fluorometric assay was used to determine entrapment and in vitro release of FITC-Ova. An HPLC method was used to determine entrapment of Quil A in PECA nanoparticles and CNP. Self-gelling microemulsion templates containing Ova and Quil A either free or incorporated in nanoparticles were subsequently investigated in an in vivo mouse model with respect to their ability to induce a sustained immune response in comparison to nanoparticles in aqueous dispersions.
Result and discussions: Biocompatible pseudoternary phase diagrams were established and characterised at 37 ˚C. Lamellar liquid crystalline gel regions and two phase turbid gel regions were found adjacent to a microemulsion region. This raised the possibility to formulate microemulsion templates converting into one phase and two phase liquid crystalline gels upon aqueous dilution. The appearance of a birefringent texture, characteristic for lamellar liquid crystals, after injection of the microemulsions into HPMC gels supported the hypothesis of an in situ phase transition. The average size of the PECA nanoparticles containing FITC-Ova was found to be in the range of 244-339 nm. The size of the nanoparticles was found to be larger in DG based microemulsions than CCG based microemulsions. The viscosity of microemulsions and liquid crystalline gels was found to show Newtonian and pseudoplastic flow behaviour, respectively. The viscosity of the microemulsion templates did not show any significant differences with increasing oil or water content, either in the presence or absence of nanoparticles. Further, the viscosity of one phase liquid crystalline gels was found to increase with increases in water content and to decrease with increases in oil content, both in the presence and absence of nanoparticles. In contrast, the viscosities of two phase liquid crystalline gels decreased with increases in water content and increased with increases in oil content. When the viscosity of one phase and two phase liquid crystalline gels were compared for systems formulated with the two different types of surfactants, the viscosity was found to be higher for DG based systems as compared to CCG based systems. The release of FITC-Ova from the selected microemulsion templates did not show any statistical differences due to the non-significant differences in their viscosity. The release of FITC-Ova from both DG and CCG based microemulsion templates was slow for the first 18 h with a rapid increase to 24 h, after which they attained a constant profile. In contrast, FITC-Ova release from one phase and two phase liquid crystalline gels was sustained for both antigen encapsulated in the nanoparticles and for free FITC-Ova. When the release from microemulsions, one phase and two phase liquid crystalline gels was compared for systems formulated with the two different types of surfactants used, release was found to be similar from microemulsion templates but was higher from CCG based one phase and two phase gels as compared to DG based gels. This was due to the lower viscosity of CCG based gels as compared to DG based gels. The entrapment of FITC-Ova in PECA nanoparticles was found to be in the range of 37.3% to 54.6% determined by a direct assay and 41.3% to 58.9% determined by an indirect assay, respectively. FITC-Ova entrapment determined by both methods was thus found to be in fairly good agreement. FITC-Ova entrapment in PECA nanoparticles however, did not follow any detectable trend with increasing oil or water content in the various formulations. The DG containing microemulsions and liquid crystalline gel formulations from the in vitro characterisation studies were selected and used to examine their ability to stimulate an effective immune response in an in vivo study. In vitro characterisation of the selected formulation was again carried out in terms of entrapment and in vitro FITC-Ova release. Additionally, entrapment of Quil A in PECA nanoparticles dispersed in the formulation was determined and found to be 46%. For comparison, a CNP dispersion in the same microemulsion formulation was prepared and entrapment of Quil A and FITC-Ova was found to be 27% and 56% repectively. The release of free FITC-Ova from the formulations was found to be faster than for encapsulated FITC-Ova for both PECA and CNP containing microemulsions. Surprisingly, a reduced T cell expansion and T cell proliferation were seen in lymph nodes and spleen after subcutaneous administration of microemulsions containing particulate formulations compared to aqueous nanoparticle dispersions. Cytokine levels, especially IFN-γ, were found either lower or comparable in mice immunised with sustained release microemulsion vaccine formulations as compared to nanoparticle dispersions. Ova specific IgG antibody titres however, showed lower antibody responses in mice immunised with particulate vaccines as compared to microemulsion based formulations, but this result cannot be further interpreted as higher response were also obtained with antigen dissolved in PBS and in the presence of alum (i.e. the control groups). Of major importance however was that the microemulsion formulations induced significant injection site reactions curtailing further animal experiments.
Conclusions: Microemulsion formulations were successfully prepared and modified to incorporate FITC-Ova into PECA nanoparticles. The microemulsions demonstrated phase transition capability to liquid crystalline gels upon absorption of small amounts of water. The nanoparticles containing microemulsions can thus be used as in situ gelling slow release formulation for antigen, encapsulated in nanoparticles as a sustained release formulation to deliver antigen for a prolonged period of time. However, an advantage of these systems to stimulate higher immune responses, compared to aqueous antigen containing nanoparticle dispersions, was not found, as the latter showed better immune responses as compared to particle containing microemulsions. The reasons for this may be that the release of encapsulated antigen from the in situ forming liquid crystal is too slow due to the slow diffusion of the nanoparticles and that the release of free antigen is also slow due to extensive protein-gel binding.