Search

Three problems concerning the operation of enzyme reactors for the recycle of ATP were investigated. The potential of a simple enzyme thermistor system for monitoring reactor effluent was studied. Initial results from physical characterisations showed the interdependence of operating parameters on overall detector sensitivity. Subsequent results demonstrated the detection of urea using an immobilized urease column. Attempts to improve the flexibility of the system by replacing covalently bound enzymes with an adsorption system proved unsuccessful. Ultrafiltration reactors were considered the most suitable type for cofactor requiring systems. The effects of various operating parameters on the problem of concentration polarisation were studied. Results indicated that transmembrane pressure drop was not linearly related to protein concentration at the membrane, as determined by the 'Gel Polarisation' model. The degree of concentration polarisation was found to be a function of temperature, tangential flow and ionic strength of the solvent. The possibility of in situ removal of denatured enzymes was considered. It was found that with suitable adjustments to the operating conditions, proteases could be profitably used for this purpose. A simple two step coupling procedure was developed for binding ATP/ADP to dextran. In the preliminary step, dextran was activated using a bis oxirane in a reaction time of 45 minutes. Subsequent binding of ATP to activated dextran was found to be time dependent allowing a range of complex loading to be produced. The resultant complex was shown to be active with a range of enzymes. Essentially all cofactor groups were shown to be available to hexokinase at the highest loading produced. The effect of cofactor loading on the apparent kinetic constants obtained with hexokinase suggests that the cofactor groups are all equally available. The potential of this complex was demonstrated in a small scale ultrafiltration reactor with hexokinase and acetate kinase.

In this thesis, glucose responsive hydrogels based on cross-linked dextran molecules were studied to determine the diffusion rate of an insulin analogue. Investigations of the interaction between concanavalin A and dextran, both in free solution and in the form of glucose responsive hydrogels were conducted. The free solution results have shown that there is an increase of association constant between concanavalin A and dextran when the molecular mass of the dextran is increased. Free solution viscometric tests have shown that increasing the molecular mass or the concentration of the dextran increases the viscosity. The hydrogels have been shown to form for dextrans of molecular mass 43kD or greater. Experiments conducted with hydrogel membranes in a diffusion cell have shown that the batch to batch reproducibility of hydrogel transport properties is low. However, clear evidence of glucose enhanced transport was obtained and these results were compared with predictions obtained from a theoretical model of gel permeability that accounts for competitive displacement of affinity cross links. Oscillatory rheological tests of gelation mixtures which showed an increase in complex viscosity at the gel point with increasing molecular mass of dextran were in agreement with empirical observations that gels formed from the highest molecular mass dextrans were more physically robust and easier to handle. Swelling rate experiments have shown that the rate of hydration of a hydrogel in the presence of glucose is decreased due to the osmotic pressure of the glucose. This work has shown that the multivalent nature of concanavalin A may not be a necessary pre-requisite for this type of hydrogel due to spatial constraints decreasing the number of potential affinity bonds per tetramer. In-house production of more tightly defined dextrans might be expected to reduce heterogeneity and improve the reproducibility of this type of hydrogel membrane.