HUNGARU\NJOURNAL OF INDUSTRIAL CHEMISTRY VESZPREM Vol. 30. pp. 87-93 (2002) KINETIC STUDY AND MATHEMATICAL MODELLING OF GROWTH OF ARTHROBACTER OXYDANS CELLS COVALENTLY IMMOBILIZED ON CELLULOSE GRANULES L.YOTOV A1, I. TSIBRANSKA, I. LALOV1, M. KARSHEVA and M. KRYSTEVA 1 (Dept. of Chemical Engineering, University of Chemical Technolugy and Metallurgy - Sofia; "Kl.Ochridsky" bul. 8, 1756 Sofia, BULGARIA 1Dept. of Biotechnology; University of Chemical Technology and Metallurgy- Sofia; "Kl.Ochridsky" bul. 8, 1756 Sofia, BULGARIA) Received: February 12,2001 The kinetics of covalently immobilized cells' growth in comparison with that of free cells was studied experimentally and numerically. Cells of Arthrobacter Oxydans were covalently immobilized to regenerated and partially deacetylated acetylcellulose granules consecutively activated with sodium periodate, urea and formaldehyde. The comparison of the data obtained showed that free and immobilized cells have the typical three phases of growth. The initial phase of free cells lasted about 10 hours and for immobilized ones ~ not more then 5 hours. The second exponential phase showed much more significant difference for both types of cells. The curve for immobilized cells is steeper than that of free oyes and attains twice faster the third phase. The glucose utilization by immobilized cells was completed about the 15th hour, for the free cells - about 25th hour. The reproducibility of the data was quite satisfactory {mean error 7.42 %). A mathematical model was used to describe and analyse the effect of the different mass transfer and reaction kinetic parameters. It shows that replication of immobilized cells is a complex process including mother cells and the new born ones. The application of the results of this study could be concerned with the use of the immobilized cells as a permanent source of free ones. They are viable and possess a long lasting capacity for urea transformation. Keywords: Arthrobacter oxydans, covalent immobilized cells of Arthrobacter oxydans, mathematical modelling of immobilized cells' growth Introduction Immobilization of microbial cells for the production of useful products has been widely studied in recent years [1]. The use of immobilized cells instead of immobilized enzymes has many advantages like avoiding the stage of isolation and purification of the enzyme, addition of co- factors and increasing the stability of enzyme system. That is the reason of the intensive development of immobilization of intact ceHs to insoluble carriers over the last years. Microbial cells can be immobilized to insoluble matrices by adsorption [2], microcapsulation [3] or covalent binding [4]. The later technique has not been widely used because of toxic nature of the reagents used for binding, causing the cells' death even during the immobilization. On the other hand the covalent binding offers advantages concerning the absence of diffusion limitations and allows the direct contact between the bound cells and the substrate solution which results in an intense (fast) controlled reaction. A new method for immobilization of proteins to cellulose derivatives was reported recently. It involves subsequent treatment of the matrix with sodium periodate, urea and formaldehyde [5] and was also applied to covalent binding of A.simplex cells [6]. The aim of the present investigation was to study the growth kinetics of covalently immobilized cells to insoluble matrices via hydroxymethyl groups and to compare it with free ones and to propose a mathematical model properly describing the process. For these studies we choose the covalently immobilized cells of Arthrobacter oxydans. 88 Materials and methods Chemicals Urea, formaldehyde, potassium hydroxide and sodium periodate were supplied by Merck (Germany); potassium dihydrogen phosphate, sodium hydrogen phosphate and glucose were obtained from Reanal (Hungary). All other chemicals were of reagent grade or better. Granules of triacetylcellulose were obtained by treatment of waste film tapes as described elsewhere [8]. Microorganisms and culture conditions Arthrobacter oxydans strain 1388 from the National collection for industrial and cell cultures (Bulgaria) was cultured on solid agar medium for 48 hours at 28°C. After this incubation colonies were picked off and suspended in meat peptone broth containing 1 g/1 sorbitol and incubated for 48 hours at 28°C. After centrifugation (20 min, 3500 g), the cell mass was suspended in nutrient medium according to Schneider and Kaltwasser [7]. The composition of the nutrient media in respect to 11 quantity was 9 g Na2HP04.12 H20, 2g KH2P04.7H20, 0.7 g MgS04.1H20, 0.02 g CaCh.2H20, Sg of glucose, 0.3 g of urea and 0.002 g NiCl2. The mineral trace elements solution for 100 ml was as follows: 0.3g FeCI3.6H20, O.lg NazMo04.2H20, 0.4g HB03, 1.34g CaC12.2H20. One ml of this solution was added to a litre of nutrient medium. The pH of the sterilized nutrient medium was 7 .0. The cultivation of free cells was done in a water bath shaker at 28°C for 48 hours. The cell mass was separated by centrifugation, washed with O.IM phosphate buffer (pH 7) and stored at 4°C. Immobilization procedure Treatment of the cellulose carrier Acetylcellulose granules with diameters in the range 1 - 3 mm were subjected to partial hydrolysis of acetyl groups according to Chen and Tsao [8]. The quantity of 0.28 g dry beads was treated with O.IM KOH and 5% (v/v) ethanol at hydromodule 1:15 for 24 hours. 1 ;te urea derivative was then prepared by oxydation of the cellulose granules with 0.25M sodium periodate at 22± 2°C at pH 5 for 2 hours in darkness. After complete removal of sodium periodate by filtration and washing the granules were treated with 15% urea solution for 14 hours in presence of 0.9% (v/v) sulphuric acid at 60°C according to Krysteva et al. [Sj. The beads were washed with water in a Buchner funnel until a neutral reaction of the rinsing water. The resulting urea derivative of the regenerated cellulose acetate beads contained 3.3 % nitrogen. This urea derivative cellulose granules were activated in 100 ml of 0.1 N phosphate buffer pH 7.5, containing 12.5% (v/v) formaldehyde and stirred for four hours at 45°C in a closed vessel. Then the granules were thoroughly washed until the complete absence of formaldehyde in the rinsing water. Immobilization of Arthrobacter oxydans The activation procedure of the carrier was followed by an immediate addition of Arthrobacter oxydans cell suspension with a concentration of 50 mg/ml. The binding was carried out at pH 8 and careful stirring at a temperature 20°C for 20 hours. The cell loaded granules were washed with water and 0.1M phosphate buffer (pH 7) until complete absence of free cells in the rinsing water, followed spectrophotometrically at 660 nm wave length. After the immobilization procedure the obtained biocatalysts granules were resuspended in the culture medium as already described and incubated for 45 hours to initiate their replication. Analytical procedures Measurement of cell growth Biomass of free cells and those, produced by immobilized cells was measured spectrophotometrically at 600 nm wave length (Perkin- Elmer spectrophotometer, Lambda 2, Germany). Cell growth was also determined by the dry cell weight according to Mallette [10]. The samples were dried until a constant weight at 105°C. The analysis were checked up by determination of protein content, using modified Lowry's method [13]. Enzyme assays Urease activity of free and immobilized cells was determined according to Melnyk and Olean [12]. The enzymatic reaction was carried out in batch vessel with magnetic stirrer. After 15 min the reaction was interrupted by cooling the mixture in 2°C bath. An aliquot (I ml) was tested for ammonia presence using Nessler reagent. Glucose determination The amount of the glucose consumed was determined colorimetricaUy at 540 nm 1ly the dinitrosalicylic acid method according to Miller [13]. ~5 r:: 0 '.;::> I!! 1: 4 0) 0 r:: 0 0 3 0) g 'Iii .0 2 ::l (/) 10 15 20 25 30 Time,(h) Fig .I a Kinetic of substrate utilization at initial concentrations for the free cells: 40.0 mg/1 (1); 118 mg/g (1 ')and for the immobilized cells: 38.4 mg/1 (2); 126mg/l (2') s ~1o 15 20 25 30 Time,(h) Fig.J b Kinetic of cells' growth at initial concentrations for the free cells: 40.0 mg!l (1); 118 mg/g (1 ')and for the immobilized cells: 38.4 mg!l (2); 126mg!l (2') Kinetic experiments The experiments on the kinetics of growth of free and immobilized cells of Arthrobacter oxydans were performed by batch process in shake flask cultures at pH=7, temperature of 28°C under continuous aeration. During the experiments the biomass and substrate concentrations were measured every 2 hours. The decrease of glucose concentration was also followed. Th~ cells of Arthrobacter oxydans were seeded in 50 em cultural medium 3 This inoculate was us~ for kin~tic studies (15 em of the inoculate seeded m 150 em of nutrient medium). The initial concentration of dry cells was varied from 40 mgll up to 118 mg/1. The initial concentration of the immobilized cells was varied in the same range as for the free ones (from 38.4 mgll to 126 mgll). 89 1.150 ~1.25 >{ .r:: ~1.00 e 0) ~0.75 .Q .00.50 10 15 20 25 30 Time,(h) Fig.2a Reproducibility of substrate utilization for free cells: first experiment (1); second experiment (1') and for im- mobilized cells: first experiment (2); second experiment (2') 10 15 20 25 30 Time,(h) Fig.2b Reproducibility of cells' growth for free cells: first experiment (1); second experiment (1 ')and for immobilized cells: first experiment (2); second experiment (2') Results and discussion Batch experiments A set of experiments for monitoring of growth of free and immobilized cells of Arthrobacter oxydans was done by analysing the biomass growth (Fig.Ja) and the substrate consumption (Fig.Jb). For both cases the three typical growth phases were observed but their duration for the free and the immobilized cells was different. As it could be seen in Fig.}, the initial phase for the free cells was about two times longer that that of the immobilized ones. This difference becomes more significant for the phase of exponential growth. The curve of immobilized cells is much steeper than that of the free ones which results in twice faster reaching of the third phase. These observations are confirmed by monitoring of substrate consumption for the two types of ceUs: the glucose utilization of the immobilized cells was completed at 90 Fig.3 Electronmicroscopic photographs of immobilized cells x 25000 about 15th hour and for the free ones - about the 25th hour. To check the reproducibility of the results obtained, parallel experiments were done. The results of two of them are given in Fig.2a and b. The obtained reproducibility of the data is quite satisfactory (mean error 7.42% ). The microscopic picture of immobilized cells of Arthrobacter oxydans is presented in Fig.3. It was found that the cells of A. Oxydans are superficially well situated on the matrix which made us suggest the absence of diffusion limitations. In the same time their covalent binding is probably limited to permit them a free reproduction. The results for the growth of young cells produced by the immobilized ones confirm this suggestion. The urease activities of free, immobilized and new born cells, produced by immobilized cells were respectively: 0.325 mol!(min.g), 0.156 moll(min.g) and 0.327 moll(min.g), based on dry cells weight. These values remained almost constant after tenfold transformations after preinoculation of the immobilized cells in a nutrient medium. These results are better than those for Escherihia coli OA5 cells with urease activity entrapped in alginate-poly-L-lysine-alginate .. microcapsules. In this case bacteria leakage has been found to occur when encapsulated bacteria were used for a four-cycle transformation [13]. In other investigations with immobilized fungi in nylon webs for enzymes production the successively repeated batch cultivations were 6 [9] or only 2 [14}. Mathematical modelling First the free cells runs were analysed using the simple Monod kinetic model: s Jl = f.l. K s + S ' in order to obtain the parameters JAm and K5• As a first approximation the experimental curves were fitted with a linear kinetics