NEW SCIENTIFIC RESULTS/ THESES

Thesis 1

Variants of precipitation - cross-linking, emulsion cross-linking, and ionotropic gelation - cross-linking methods have been developed to obtain chitosan support particles applicable for immobilization of β-D-galactosidase, in form of macrospheres, microspheres and nanoparticles, respectively. Among them, chitosan microspheres were considered as the most suitable support.

1.1. The precipitation - glutaraldehyde cross-linking method resulted in smooth, non porous chitosan macrospheres of 3.2 mm mean diameter with narrow size distribution. It was found that upon functionalization of macrospheres in order to couple the enzyme with 1% and 5% glutaraldehyde, an increase of hardness and rigidity of particles occurred.

1.2. To prepare chitosan microspheres three different methods were investigated:

precipitation by spraying chitosan solution into a stirred precipitating agent, ionotropic gelation under mixing, and emulsion linking. It was found that emulsion cross-linking was the most appropriate technique, producing narrow size distribution and non-aggregated particles. By proper selection of the experimental parameters, such as oil phase composition, nature and concentration of surfactant, a wide range of mean particle size between 20 and 500 µm was obtained. The petroleum ether - paraffin oil mixture, recommended by other authors, was replaced with more ecofriendly nhexadecane -sunflower oil mixture. In respect of morphological properties of the obtained microparticles and suitability for enzyme immobilization, the oil phase composed of 60%

n-hexadecane and 40% sunflower oil, and Tween 80 surfactant in 3% concentration was concluded as optimal conditions.

1.3. For preparation of nanometer size support particles ionotropic gelation method followed by glutaraldehyde cross-linking proved to be appropriate. It was found that, depending on the type of gelation agent different mean sizes can be obtained: sodium tripolyphosphate and sodium sulphate resulted in 152 nm and 517 nm mean particle sizes, respectively.

(Related publications: [1], [4], [5], [9])

Thesis 2

The particular effects of process variables on the mean particle size of chitosan microspheres were studied by a 4-factorial 3-level Box-Behnken-type experimental design and statistical analysis. Stirring rate, Tween 80 emulsifier concentration, chitosan concentration and glutaraldehyde concentration were selected as parameters with significant influence on the resulted mean particle size.

2.1. It was found that at high concentration of cross-linking agent increasing stirring rate diminished the achievable mean particle size, due to the enhanced dispersion effect.

However, at low cross-linking agent concentration stirring had opposite effect, supposedly due to the higher susceptibility of the not fully solidified particles for coalescence.

2.2. The increase of surfactant concentration generally decreased the mean particle size, but at higher cross-linking agent concentration this effect disappeared, probably due to the higher viscosity or faster solidification of droplets resisting more against their break up.

2.3. It was found that the concentration of chitosan solution had the most important effect, and generally the higher its value the smaller the particles formed by emulsion cross-linking, probably due to the faster solidification and decreasing susceptibility of droplets for coalescence at higher chitosan concentrations. Less coalescence after the initial break up of the droplets leads to smaller particle size.

2.4. The concentration of glutaraldehyde was found as the second most important influence on the mean particle size, due to the hardness or less stickiness of particles during solidification, which was therefore closely correlated with the values of other parameters.

At high stirring rate the decrease of glutaraldehyde concentration resulted in steep increase in the mean particle size due to the higher coalescence.

2.5. As a result of statistical analysis of experimental data and non-linear regression, equation was proposed to predict the mean particle size as a function of the studied process variables with reasonable confidence and acceptable mean error. The equation proved to be applicable to optimize the process variables to obtain chitosan microspheres for biocatalyst support with proper mean particle size.

(Related publications: [2], [10], [11])

Thesis 3

On the basis of immobilization studies, it was stated that chitosan microspheres were the most appropriate potential carriers for covalent immobilization of β-D-galactosidase, using glutaraldehyde as cross-linking agent. By investigation of the main factors that influence the immobilization, the optimal process parameters were found:

3.1. Glutaraldehyde concentrations of 2% and 3% resulted in similar immobilization efficiency, but the latter concentration (3%) was considered as optimal one ensuring more coupling positions on the particles, because of the smaller size and higher specific surface of particles.

3.2. The binding capacity of chitosan microparticles cross-linked with 3% glutaraldehyde was 27 mg protein/g dry support.

3.3. It was found that the proper reaction time for coupling of enzyme to the chitosan microparticle was 6 h, ensuring a maximal value of total activity yield.

3.4. In the most favourable conditions the immobilization yield of protein reached 99% at 3.7 mg loaded protein per 1 g wet chitosan beads and the recovery yield of total enzymatic activity following immobilization was 13.06%. The reduction of Schiff bases with sodium borohydride after the covalent attachment, providing more flexible and stable secondary amino bonds, has increased the catalytic efficiency almost twofold resulting in a total activity yield of 23.5% which is a reasonable result for covalent immobilization.

(Related publications: [1], [3], [8], [12])

Thesis 4

β-D-Galactosidase has been immobilized by sol-gel entrapment in organic-inorganic composite matrices containing covalently linked alkyl groups or using combined sol-gel entrapment and adsorption. This study resulted in nanostructured biocatalysts with high immobilization yield and activity.

4.1. Among several silane precursors evaluated, tetraethoxysilane and methyl-triethoxysilane at 7:1 molar ratio led to the highest activity of 3252.78 µmol ONP×min^-1×g^-1 of immobilized β-D-galactosidase at protein loading value of 6.22 mg per mmol silane precursor.

4.2. The maximal immobilization efficiency was obtained at lower protein loading of 4.1 mg/mmol silanes, when the recovery yield of total activity was more than 400%. Such a high yield demonstrates that enzyme activity has increased after encapsulation in the nanostructured sol-gel matrix. The possible reasons are that the enzyme is evenly distributed inside the porous matrix preventing aggregation of the biomolecules, and the microenvironment inside the matrix protects the enzyme against inactivating effects.

4.3. Sol-gel entrapment combined with adsorption resulted in the best catalytic efficiency when chitosan was used as adsorbent. The obtained activity was very close to that of β-D -galactosidase entrapped in simple sol-gel matrix (about 3,300 enzymatic units/g support).

Thesis 5

Immobilized β-D-galactosidase biocatalysts obtained by both covalent coupling and sol-gel entrapment have been characterized, and their pH, temperature, operational and storage stabilities were determined.

5.1. It was found that the covalently attached β-D-galactosidase was slightly more stable at higher pH values compared to the native enzyme (showing about 50% residual activity at pH 9.0), while the sol-gel entrapped enzyme exhibited excellent stability at alkaline pH values up to pH 11 (about 90% residual activity, compared to the maximum value at pH 8.5).

5.2. The sol-gel entrapped β-D-galactosidase showed markedly improved thermal stability compared to the native and covalently immobilized enzymes. Inactivation was noticed only at 60°C after 8 h incubation, whereas the native enzyme lost about 50% of activity after 3 h at 40°C.

5.3. The Michaelis constants (K_m) for lactose hydrolysis determined by kinetic studies were 140.7 mM for the covalently bonded, 123.7 mM for the sol-gel entrapped β-D -galactosidase and 84.4 mM for the non-immobilized enzyme. The decrease of specificity towards lactose was probably due to the higher mass transfer resistance and diffusional limitation that is usual in employing immobilized enzymes.

5.4. It was found that the sol-gel entrapped enzyme demonstrated high storage and acceptable operational stability. After 5 reuse cycles, the residual activity was still about 50% of the initial value.