Characterization of. kappa.-carrageenan gels used for immobilization

ner, 1983; Karel et al., 1985; Chao et al., 1986). The polymer-cell mixture ... NH4+) is necessary to ensure structural integrity of K-car- rageenan g...
0 downloads 0 Views 3MB Size
Biotechnol. Prog. 1991, 7, 516-525

516

Characterization of K-Carrageenan Gels Used for Immobilization of Bacillus firmus Seung-Hyeon Moon and Satish J. Parulekar* Department of Chemical Engineering, Illinois Institute of Technology, Chicago, Illinois 60616

In this study, aimed a t a biochemical and physical characterization of K-carrageenan gels used for entrapment of Bacillus firmus NRS 783 (a superior producer of an alkaline protease), effects of carrageenan concentration, gelation temperature, initial cell loading, and strength of the curing agent (KC1) on the properties of cell-free and cell-laden gels were examined. The physical properties of the differently prepared gels that were examined included density, free volume fraction, mechanical strength, and change in gel volume during gel curing. The biochemical characteristics studied included viability of gel-entrapped cells, cell leakage from cell-laden gels, and cell penetration into cellfree gels. For the range of carrageenan contents investigated [between 2 % and 5 % (w/v)], the mechanical strength of the gels with/without KC1 curing was observed to increase with a n increase in carrageenan content of gels. The mechanical strength of each gel increased substantially upon extensive curing. Free volume fractions in excess of 0.8 were observed for all gels. Of cells that were viable prior to immobilization, 90-92 % remained viable after formation and extensive curing of gels for cell-gel mixtures prepared at 45 "C. Attempts at prolonged storage of cell-laden gel beads a t 0 "C as stock cultures of immobilized B. firmus were unsuccessful due to a significant decline in cell viability during such storage. On the basis of the cell leakage studies, the average pore sizes of 296, 3 % , 496,and 5 % gels were deduced to increase in the following order of carrageenan content (w/v): 4% , 3 % , 25% , and 5 % . Commensurate with the decrease in the average pore size (or the increased tightness of the gels) with increasing carrageenan content, both the extent of cell leakage and the extent of net cell penetration decreased with increasing carrageenan content for the first three gels. Owing to nonuniform distribution of free space and much larger pores, the extent of net cell penetration in 5 % carrageenan gels was considerably low, while the extent of cell leakage in 5 % carrageenan gels was an order of magnitude greater than the extents of cell leakage in the other three gels.

Introduction Entrapment within a porous matrix protects the cells from the shear field outside the support particles and enables high retention of cell viability (Chibata and Tosa, 1981; Chibata et al., 1983; Karel et al., 1985; Woodward, 1985). The porous matrices used for immobilization are either preformed or formed in situ. Physical entrapment of organisms inside a polymeric matrix formed in situ (gel) is one of the most widely used methods for whole-cell immobilization. The gels commonly used for cell entrapment include agar, alginate, K-carrageenan, and polyacrylamide. Gel entrapment is usually based on suspending the cells to be immobilized in a solution containing watersoluble polymer(s), followed by induction of gelation of the polymer-cell mixture via physical or chemical means (Takataet al., 1982;Birnbaum et al., 1983;Klein and Wagner, 1983; Karel et al., 1985; Chao et al., 1986). The polymer-cell mixture may be either gelled immediately into a final desired shape and size or gelled in sheets and then cut into particles of the desired dimensions. Despite considerable research on whole-cell immobilization (Chibata and Tosa, 1981;Chibata et al., 1983;Karel et al., 1985;Woodward, 1985),the proper design of support particles with desirable cell-support interactions remains inadequately addressed. The procedures followed for

* To whom correspondence should be addressed. 8756-7938/91/3007-0516$02.50/0

support preparation (formation and curing in the case of gels) can have a significant impact on the properties of the immobilized cell-support particle aggregates and hence on the performance of immobilized cell processes. In several previous studies, "preassigned" procedures for support preparation have been followed without proper justification as to the suitability of these for a particular biochemical system. With P-D-galactose sulfate and 3,6-anhydro-a-~-galactose as its building blocks, K-carrageenan has molecular weight in the range 100 000-800 000 (Tosa et al., 1979). Dissolution of K-carrageenan in water at temperatures equal to or in excess of 70 "C and cooling of the resulting mixture in the presence of cations such as K+ and Rb+ leads to formation of stable carrageenan gels (Chanzy and Vuong, 19851,which are suitable for cell entrapment. The thermally reversible sol-gel transition is accompanied by a cooperative coil-helix transition that involves chain dimerization, i.e., double helix formation (O'Neill, 1955; Reid et al., 1974; Rees, 1975; Norton et al., 1978; Morris et al., 1980;Dea, 1981;Borchard, 1983). The double helix structure has been characterized in the condensed (gel) phase by X-ray fiber diffraction (Anderson et al., 1969; Arnott et al., 1974;Rees, 1972)and shown to persist under conditions of extensive hydration by optical activity and NMR relaxation studies (McKinnon et al., 1969; Rees et al., 1969; Bryce et al., 1974). Helices appear to form and are stable only under conditions that also promote helix-

0 199 1 American Chemical Society and American Instiiute of Chemical Engineers

517

Biotechml. hog., 1991, Vol. 7, No. 6

helix aggregation (Morris et al., 1980; Rees et al., 1982). The bulk properties of carrageenan gels are determined by quaternary interactions between double helices, in addition to direct interchain association through doublehelical junction zones (Rees et d., 1982). Owing to the polyanionic (sulfate) nature of K-carrageenan, the presence of cations (such as Rb+, K+, Cs+,and NH4+) is necessary to ensure structural integrity of K-CLUrageenan gels (Rees et al., 1982). At comparable ionic concentrations, the strongest gels are obtained with K+ and Rb+, significantly weaker gels are obtained with Cs+ and NH4+, and formation of gels with cohesive structure is not possible with Li+ and Na+ except a t very high concentrations of these two ionic species (Morris et al., 1980; Rinaudo and Rochas, 1981; Rees et al., 1982). For each cationic species, there exists a critical ionic concentration above which gel formation can occur. For concentrations of the cationic species in excess of the critical concentration, temperatures corresponding to sol-gel and gel-sol transitions are different. In the case of K-carrageenan gels, the critical Na+ concentration is severalfold (nearly 28 times) higher than the critical K+concentration (Rinaudo and Rochas, 1981). At their respective concentrations employed in the present study, gel stability is assured in the presence of K+ and exposure to Na+ (in the absence of K+) results in disintegration of K-carrageenen gels. The effects of (i) carrageenan concentration, (ii) gelation temperature, (iii) initial cell loading, and (iv) strength of the curing agent (KC1) on properties of cell-free and cell-laden gels were examined in the present study. The physical properties examined included (i) density, (ii) free volume fraction, (iii) mechanical strength, and (iv) change in gel volume during gel curing. In a nongrowth environment, the following biochemical characteristics (indicators of the nature of cell-support interactions) were studied for Bacillus firmus NRS 783: (i) cell viability, (ii) cell leakage from cell-laden gels, and (iii) cell penetration into cell-freegels. B. firmusNRS 783is a superior producer (on an industrial scale) of an alkaline protease (Aunstrup et al., 1974; Priest, 1982; Helmo et al., 1985; Kalisz, 1988; Moon and Parulekar, 1991).

Materials and Methods Preparation of Cell-Free and Cell-Laden Gels. K-Carrageenan (Sigma Chemical Co.) was dissolved in distilled water a t 70 "C and the resulting solutions were autoclaved for 15 min. For preparation of cell-laden gels, carrageenan solutions were cooled to 45 "C and mixed with appropriate volumes of cell broth, the resulting mixtures being stirred immediately. The carrageenan solutions/suspensions, poured into petri dishes or shake flasks, were cooled to room temperature (25 "C) for formation of gels. Gel plates of desired dimensions were obtained by cutting the hardened gels on petri dishes by a sterile razor blade. The gel plates were soaked in a 2.1 5% (w/v) KC1 solution for a t least 24 h. For preparation of spherical gel beads, carrageenan solutions/suspensions kept at 45 "C were pumped dropwise through a hypodermic needle (no. 18G11/2, Becton Dickinson, Rutherford, NJ) into a 2.1 % (w/v) KCl solution kept at room temperature (25 "C). Densities and Free Volume Fractions of Gels. The densities of cell-free gel cubes (around 1cm X 1cm X 1 cm in dimensions) subjected to KC1 curing for 24 h or more were obtained by measuring the mass and volume of these. The dependence of volume of gel particles, if any, on the strength of the curing solution was estimated

by transferring cell-free gel cubes (around 1cm X 1cm X 1cm in dimensions), cured for 24 h with a 2.1 % (w/v) KC1 solution, to KC1 solutions of various strengths [&lo% (w/v)]. The change in gel volume was obtained by measuring the gel volumes before transfer to a KC1solution of strength other than 2.1% (w/v) and after transfer to and treatment with the KC1 solution for 15h or more. The diameters of spherical gel beads were measured precisely by Vernier calipers. To estimate the free volume fraction of cell-free gels, gel beads [cured with a 2.1% (w/v) KC1 solution] of a known volume (about 3 mL) were transferred to 250 mL of an aqueous solution of glucose (10 g/L) and KC1 [2.1% (w/v)]. The presence of KC1 was necessary to prevent swelling and disintegration of gel beads during their exposure to glucose. After glucose was allowed to penetrate into gel beads for 24 h or more so that an equilibrium with respect to glucose was established between the beads and the solution, the beads were removed from the solution and disintegrated in a 0.8% (w/v) NaCl solution a t 45 "C. The amount of glucose penetrated in the beads was estimated by measuring the volume of the resulting solution and concentration of glucose in it. Glucose concentrations in this resulting solution and the bulk glucoseKC1solution left over after withdrawal of gel beads were measured using a Sigma enzymatic glucose assay kit. Since the volume of the glucose-KC1 solution was much greater than that of the gel beads, changes in glucose concentration in this solution during these experiments were observed to be insignificant. The free volume fraction of the cell-free gels was then calculated as free volume fraction =

-(

=1

volume of porous space in gel total volume of gel

amt of glucose penetrated in gel beads

K glucose concn in the glucose-KC1 solution) volume of gel beads The distribution or partition coefficients for glucose, K [(g of glucose/L of porous space in gel)/(g of glucose/L of bulk solution)], for various gels were estimated as follows. Cell-free carrageenan solutions were allowed to set a t 0 "C for 24 h on petri dishes to form gels of height 1cm. Gel cubes (approximately 1cm X 1cm X 1cm in dimensions) cut from these gels were cured in a 2.1 % (w/v) KC1solution for more than 24 h. The cured gel cubes were then transferred to an aqueous solution containing glucose (10 g/L) and KC1 [2.1% (w/v)] as described above. After allowing contact for 24 h or more between gel cubes and glucose-KCl solution, the gel cubes were removed from the solution. After the external surface of gel cubes was dried with a filter paper (Whatman #l), the gel cubes were placed on petri dishes. The dried gel cubes were pressed slightly (at room temperature) to release the solution in their porous space while avoiding their rupture. The partition coefficient of glucose for each gel was then obtained as the ratio of glucose concentration in the solution leaked from a gel to glucose concentration in the glucose-KC1 solution. Mechanical Strength of Cell-FreeGels. The critical pressure required to rupture a gel and the critical tensile stress required to fracture a gel were measured as follows. Cell-free carrageenan solutions were allowed to set a t 0 "C for 24 h on petri dishes to form gels of height = 6.6 f 0.4 mm. Gel plates of two types, one with near-square cross section [approximately 1.25cm X 1.25 cm in cross-sectional area and 6.6 f 0.4 mm in height, the dimensions of the

518

Biotechnol. Prog., 1991, Vol. 7, No. 6 Fixed Plastic Casing

Movable Plastic Casing

Spring Force

Figure 1. Schematics of the device used for measurement of elastic or tensile strengths of K-carrageenan gels.

plates being measured by Vernier calipers] and the other with I-shaped cross section (6.6 f 0.4 mm in height, the smaller dimensionof the neck of the I-shaped cross-section being 1 cm), cut from these gels were cured in a 2.1% (w/v) KC1 solution for more than 24 h. The critical pressure required to rupture the gel plates with near-square cross-section at room temperature was measured using a Bimba pressure cylinder (Model BF-042-D, Bimba Co., Monee, IL) supplied with compressed air through a pressure regulator. The air pressure was increased at the rate of 50.65 kPa/s manually until the gels ruptured. The schematicsof the device employed for measurement of tensile strengths of differently prepared gels are provided in Figure 1. Gel plates with I-shaped cross section and 6.6 f 0.4 mm in height were held between two plastic casings. A previously calibrated spring (as concerns the relationship between extension of the spring and the force applied on the spring) was attached to one of the casings. Stretching of each gel plate was accomplished by application of external force on the spring while keeping the location of the other casing fixed. The external force was increased at the rate of 0.1 N/s manually until the gel plate fractured, the fracture occurring in the neck region of the I-shaped gel plate. The tensile strength of each gel was calculated as the ratio of the minimum force necessary for gel fracturing to the cross-sectional area at the location of fracture (0.66 cm X 1 cm). Bacterial Strain, Media, and Cultivation. Stock cultures of B. firmus NRS 783 were maintained at 0 "C on the nutrient medium containing glucose (1.0 g/L), yeast extract (5.0 g/L), and tryptone (5.0g/L) with the pH being maintained at 9.7 with 0.1 M sodium sesquicarbonate buffer (Moon and Parulekar, 1991). The subculturings for preparation of cell broth required for formulation of cell-laden gels and cell penetration studies were carried out at 37 "C and an initial pH of 9.7 for each subculturing. The nutrient medium used for the first subculturing was that used for preparation and maintenance of stock cultures, while a semidefined medium (Moon and Parulekar, 1991) was used for the second subculturing, the semidefined medium containing (per liter): (i) glucose, 6 g, and yeast extract, 0.3 g; (ii) K2HP04,2 g; and (iii) NH4C1,4 g; MgS04, 0.7 g; FeSOq7H20,0.02 g; ZnS04.7H20, 0.024 g; MnC1~4H20,0.04g; CaC1~2H20,0.06g; NazMo04, 0.001 g; nitrilotriacetic acid, 0.35 g; citric acidoH20, 0.017 g; CuC1~2H20,0.003 g; CoC1~6H20,0.008 g; and Na2B40~10H20,0.002g, with the three parts [ (i-iii)] being autoclaved separately. For preparation of cell broth for cell penetration studies, the semidefined medium was

supplemented with 2.1% (w/v) KC1 to prevent disintegration of gel beads during cell penetration experiments. Cell Viability. Gel beads containing a known number of viable B. firmus cells and cured with a 2.1 % (w/v) KC1 solution for at least 24 h were disintegrated in a 0.8% (w/v) NaCl solution at 45 "C. The volume of the saline solution was approximately 20 times the volume of gel beads. After the volume of the resulting mixture was measured, the concentration of viable cells in it was measured by colony counts of an appropriately diluted portion of the mixture on nutrient-agar plates containing 5 g/L tryptone, 5 g/L yeast extract, 1g/L glucose, 15g/L agar, and 0.1 M sodium sesquicarbonate buffer. Cell Leakage. For estimating the extent of cell leakage, carrageenan suspensions (50 mL) containing a known number of viable B. firmus cells were prepared at 45 "C in 500-mL shake flasks, allowed to set by cooling to room temperature, and cured by contacting with a 2.1 % (w/v) KC1 solution for 24 h. The gels therefore assumed the shape of conical discs. The hardening solution was then replaced by 50 mL of a saline solution containing 0.8% (w/v) NaCl and 2.1 % (w/v) KC1. The flasks were shaken at 250 rpm and 37 "C on a water bath shaker. The presence of KCl in the saline solution was necessary to enable leakage of cells from the gel without its swelling and disintegration. Liquid samples (0.1 mL) were withdrawn periodically from the shake flasks for estimation of viable cell concentration in the saline solution, the cell concentration being obtained via colony counts of appropriately diluted samples on nutrient-agar plates. The extent of cell leakage during extended gel curing was estimated by contacting cell-laden gels with a 2.1% (w/v) KC1 solution and measuring the concentration of cells effused into the KC1 solution after extended curing (for 24 h or more). Cell Penetration. The extent of cell penetration in cell-free gel beads was measured by transferring 40 mL of cell-free gel beads to 100 mL of a cell broth kept in a nongrowth environment (a batch culture of B. firmus NRS 783 near the termination of growth phase) and agitating the resulting suspension on a shaker at 250 rpm and 37 "C. A small number of beads (10) was periodically withdrawn, washed with a saline [0.8% (w/v) NaCl] solution to remove cells attached to the external surface of gel beads, and disintegrated in the saline solution at 45 "C. Cells from an appropriately diluted portion of the resulting solution were grown on nutrient-agar plates for colony counts.

Results and Discussion Good mechanical strength and higher density (in comparison to that of nutrient medium) of gels are desirable for increased operational stabilityof immobilized cell reactors. As much as possible, the gels should be able to withstand stresses due to shear field outside the support particles, gel compression in bioreactors densely packed with gel particles, and cell growth in the porous regions of the particles. While some difference in the densities of gel particles and nutrient medium is desired for confinement of the particles to the reactor when these are freely suspended, the difference should not be so much that flotation or fluidization of the support particles is difficult. High retention of cell viability and high porosity are of interest for enhancing productivity of an immobilizedcell process. While a gel should be as highly porous as possible, the pores should be neither too small nor too large and should be uniform in size and well-distributed in the gel. Cell-free and cell-laden gels with carrageenan contents in the range 2 % -5 % (w/v) were investigated in this study.

510

Biotechnol. Prog., 1991, Vol. 7, No. 6 i 1.035

s

s

-

2 0.9

2.5 3.0 3.15 4.b 4.5 5.bl*015 Carrageenan content, percent (w/v)

0'72.b

Figure 2. Variationsin (i) density (0) and free volume fraction (0) of cell-freeK-carrageenan gels prepared at 45 "C and (ii)free volume fraction (M) of cell-free.+carrageenangels prepared at 70 "C with variation in carrageenan content of the gels. The gels were cured for 24 h at room temperature (25 "C) in a 2.1 ?' 6 (w/v) KCl solution. In this and other figures, the carrageenan content of both cell-free and cell-laden gels, as appropriate, is expressed as concentration of K-carrageenan [ % (w/v)] in carrageenan solutions/suspensionsprepared at 45 "C (unlessstated otherwise) for formation of gels.

Gels containing less than 2% (w/v) K-carrageenan were found to be very brittle and therefore are not suitable as support matrices for cell immobilization. Formation of gels with carrageenan content in excess of 5% (w/v) was not possible at temperatures less than or equal to 55 "C. Temperatures equal to or in excess of 55 "C were found to be detrimental for survival ofB. firmus,hence the upper limit on carrageenan content for preparation of cell-laden gels. Densities and Free Volume Fractions of Gels. Variation in density of K-carrageenan gels, cured with a 2.1 5% (w/v) KC1 solution, with variation in carrageenan content of gels is depicted in Figure 2. While the gel density increases with increasing carrageenan content in the range investigated [2.5%-5% (w/v)],the variation in gel density is not significant. In a previous study, Wang and Hettwer (1982) have reported the density of 4% (w/v) K-carrageenan gels treated with a 2 % (w/v) KC1solution to be 1.03 g/mL, a value very close to that observed for 4% w a r rageenan gels in the present study. The gel bead size was observed to be mildly dependent on carrageenan content of the gels, increasing with increasing polymer concentration. The average diameters of gel beads prepared at 45 "C from carrageenan solutions containing 2 % , 3% , 4 % ,and 5 % (w/v) K-carrageenanwere, for example, 3.24,3.46,3.6,and 3.9 mm, respectively; each value represents the average of diameter measurements for at least 20 beads. The diameter of a liquid drop falling in air is influenced by density, surface tension, and viscosity of the liquid (Harkins and Brown, 1919;Perry et al., 1984). Since the variation in gel density with variation in carrageenan content of gels is not significant, the bead diameter is mainly influenced by surface tension and viscosity of carrageenan solution at the gelation temperature. When carrageenan solutions were prepared at 70 "C and bead formation was initiated at this temperature, the average bead diameters of 4 % , 4.5%,and 5% (w/v) K-carrageenan gels were observed to be 3.45,3.51,and 3.58 mm, respectively. For fixed carrageenan content, the reduction in the bead diameter with an increase in gelation temperature is attributable to the accompanying decreases in surface tension and viscosity of the carrageenan solution (Perry et al., 1984). The partition coefficients of glucose (the principal carbon source in the medium used for cultivation of B. firmus NRS 783) in 2%, 3 % , 4% and 5% (w/v) carrageenan gels in the presence of KC1 were found to be 0.99,

0.98,1.03, and 1.02, respectively. Partitioning of glucose between the aqueous phases in gel beads and outside the gel beads was therefore negligible. The free volume fractions of gels formed from solutions with different K-carrageenan concentrations [2%-5% (w/v)] at 45 O C are presented in Figure 2. While an increase in carrageenan content up to 3.5% (w/v) led to a decrease in the free volume fraction of the gels, an increase in the polymer concentration beyond 4% (w/v) resulted in an increase in the free volume fraction of the gels. Gels with carrageenan content between 3.5 % and 4% had the lowest free volume fraction. A gelation temperature of 45 "C has been recommended for formation of 4 % (w/v) K-carrageenangels(Rees, 1972). The ideal gelation temperature is expected to increase with increasing carrageenan content of gels. The ideal gelation temperature for gels with carrageenan content in excess of 4% (w/v) is thus greater than 45 "C. Conduct of the bead preparation procedure at a temperature (45 "C) lower than the ideal gelation temperature is believed to have seriously affected the tightness of cell-free and cell-laden gels prepared from solutions/suspensions with carrageenan contents of 4.5% and 5% (w/v) (possibly due to aggregationof polymer chain segments) and may explain the increase in free volume fraction observed in Figure 2 with increase in carrageenan content beyond 4%.(w/v). A further corroboration for the observations in Figure 2 and this reasoning is provided by estimations of free volume fractions of gels prepared at a higher temperature. The free volume fractions of cell-free gels prepared from carrageenan solutions kept at 70 "C and with carrageenan contents of 476, 4.576, and 5% (w/v) were estimated to be 0.765,0.75, and 0.738, respectively (see Figure 2), and exhibited a trend similar to that exhibited by free volume fractions of cell-free gels prepared at 45 "C and with carrageenan contents less than 3.5% (w/v). The porosity of gels is thus strongly influenced by the temperature at which gel preparation is carried out. A further evidence for structural differences between gels prepared from 2 % , 3 % ,and 4 % (w/v) carrageenan solutions/suspensions and those prepared from 5 9% (w/v> carrageenan solutions/ suspensions at 45 "C is provided later by optical micrographs of slices of cell-laden gels. Change in Gel Volume Due to Swelling and Curing. Since the hydrogel cannot dissolve in water, it appears to take up and retain significant amounts of water. During swelling, the volume of a hydrogel increases until an equilibrium, at which the chemical potentials of water in the gel and of water surrounding the gel are balanced, is reached. The polymer chains in the gel are elongated with swelling of the hydrogel and exert a force in opposition to swelling. The osmotic pressure attributed to the polymer ( A ) is the driving force for swelling, which is counteracted by the elastic contractability of the stretched polymer network @). A t equilibrium, ?r equals p . Since swelling depends on the osmotic pressure and therefore on the chemical potential of water in solution outside the hydrogel, the degree of swelling depends on the types and concentrations of solutes in the solution. Solutes such as glucose, NaC1, and urea are known to interact with the macromolecular polymer network as polymerization inhibitors (Kalal, 1983). When K-carrageenan gel beads were immersed in a 0.1 % (w/v) glucose solution without KC1 a t room temperature, the beads swelled considerably during the first 6 h and disintegrated completely after 6 h. In the case of polysaccharide gels, the cross-links, which hold the three-dimensional network of polysaccharide macromolecules together, are usually

Biotechnol. Prog., 1991, Vol. 7, No. 6

520

-3

1

e

4*0 2.0

3

e, M

c 0.0

."

a!

6n

3 -2.0

c u -4.0

t'

0

I

I

I

I

2

4

6

8

10

KCI concentration, percent (w/v)

Figure 3. Percentile change in volume of gel cubes, cured for 24 hat roomtemperature(25OC)witha2.1% (w/v)KClsolution, upon transfer to and extensive curing (for 15 h or more) at room temperature with KCl solutions of various strengths. The carrageenan content of the gels was 3% (w/v). semipermanent as they originate either from ionic or from intermolecular associations. Treatment of gels with solutions containing cations such as K+ and Rb+ results in curing of gels (Borchard, 1983). The degree of swelling should decrease with increased cross-linking of the macromolecular network such as that resulting from more extensive curing with a hardening agent. The percentile changes in volume of warrageenan gel particles, prepared from a 3% (w/v) carrageenan solution and cured for 24 h with a 2.1 % (w/v) KCl solution, upon transfer to and further extensive curing with KC1 solutions of various strengths are depicted in Figure 3. Transfer of gel particles to solutions containing KC1 at concentrations less than 2.1 % (w/v) led to swelling of the particles. Transfer to solutions containing KCl at concentrations greater than 2.1 % (w/v) led, on the other hand, to shrinking of the gel particles. Transfer of gel beads from a 2.1% (w/v) KC1 solution to pure water led to a 16% increase in gel volume (not shown in Figure 3) but did not result in their disintegration upon exposure to water for extended periods. The decline in the percentile volume change, which is proportional to the ultimate gel volume, with increasing KC1 concentration in the curing solution illustrates that the degree of gel swelling decreases with increased cross-linking of the macromolecular network in these hydrogels. Significant changes in support particle volume can occur in operations of immobilized cell reactors (Shiotani and Yamane, 1981). In a growth environment, the average size of gel beads, or alternately the extent of swelling, depends on the extent of cell growth and the composition of the nutrient medium outside the gel beads, both of which are in general subject to spatial and temporal variations. The results in Figure 3 suggest that the osmotic pressure of the nutrient medium outside the gel beads must be properly adjusted by continual presence of the curing agent (KC1)to maintain the integrity of immobilized cell-support particle aggregates and to minimize spatial and temporal variations in size of gel beads in the immobilized cell reactor. Mechanical Strength of Gels. The polymeric support used for cell entrapment is subject to stresses due to compression, tension (stretching), and shear in operation of bioreactors. While shear stress due to fluid motion outside the support may lead to partial disintegration of support, stresses due to compression and tension can result in total disintegration of support. Significant compression of gel beads can occur (i) in bioreactors densely packed with gel beads owing to significant particle-particle

Carrageenan content, percent (w/v)

Figure 4. Variations in (i) the critical pressure required for gel rupture (A,0) and (ii) the critical tensile stress required for gel fracture ( 0 , 0)at room temperature (25 O C ) with variation in carrageenan content of gels. The symbols 0 and 0 refer to experimental data from gels cured with KCl [2.1% (w/v)] at room temperature for 24 h or more, and the symbols A and 0 refer to data from gels not cured with KC1. interactions and (ii) in the bottom portion of tall vertical bioreactors due to significant static head of liquid. Excessive cell growth in gels results at first in substantial swelling of gels due to enlargement of pores and can eventually result in disintegration of gels. The critical compressive and tensile stresses required for rupture and fracture, respectively, of gels are therefore good indicators of mechanical strength of gels. Variations in (i) the critical pressure required for gel rupture and (ii) the critical tensile stress required for gel fracture with variation in carrageenan content of gels are illustrated in Figure 4. Each data point in Figure 4 represents the average of measurements for ten replicates of gel plates, the measurements lying within &5% of the average. The higher the water content of the hydrogels, the lower is their mechanicalstrength. The critical stresses required for gel rupture and gel fracture increased, therefore, with increasing carrageenan content (or decreasing water content) of gels. The cross-linkingin gels produced upon coolingsolutions of water-soluble polymers is due to hydrogen bonding (Borchard, 1983;Finch, 1983). The intermolecular ionic bonds formed by addition of an appropriate cation (K+ in the case of K-carrageenan) are much stronger than the intermolecular hydrogen bonds. Curing of gels is therefore expected to increase their strength owing to alteration in both the extent and the nature of cross-linking. Over the range of carrageenan contents studied, the critical pressure for rupturing a gel and the critical tensile stress for fracturing a gel increased by nearly 200 % upon extensive curing of the gels with a 2.1% (w/v) KC1 solution (see Figure 4). Retention of Cell Viability upon Immobilization. The materials and methods used for immobilization may themselves cause damage to the cells to a lesser or greater extent (Koshcheyenko et al., 1983;Starostina et al., 1983). In the present study, retention of cell viability was examined in gels with various carrageenan contents after they were cured with a 2.1% (w/v) KC1 solution. The concentration of viable B. firmus cells in each cell-carrageenan suspension (prepared at 45 "C) before formation of gel beads was 1.11 X lo8 cells/mL of suspension. For carrageenan concentrations investigated [2 % -5 5% (w/v)], 90-92 % of cells that were viable prior to immobilization remained viable after formation and extensive curing of gels. The loss of a portion of viable cell population is attributable mainly to contact of cells with cell-free car-

521

Biotechnol. Prog., 1991, Vol. 7, No. 6

Table I. Retention of Cell Viability* for 3% (w/v) rr-Carrageenan Gel Beads Stored a t 0 OC storage time viability in viability in (days) KCl solutionb semidefined mediumc 1 92.0 95.0 2 92.5 3 88.0 5 43.4 86.7 I 1.4 24.2 9 a Viability is expressed as percent of the initial viable cells (in the cell suspension prior to gel formation) remaining viable. b 2.1 % (w/ v) KCl solution. Moon and Parulekar (1991).

rageenan solution at 45 "C and partly to effusion of cells from gel beads during KC1 treatment. Effusion of cells during gel curing with KC1 solution is discussed later. Preparation of cell-carrageenan suspensions a t or in excess of 55 "C resulted in complete loss of viability of B. firmus. It is desirable that immobilized cells have as long an operating life as possible. In this regard, it is of interest to explore if viability of the gel-entrapped cells could be maintained by storing the gel beads carrying these at low temperatures over extended periods. Unless there is a drastic decline in cell viability during prolonged storage of gel beads, the beads containing entrapped cells could be used as stock cultures of immobilized cells. With this intention, 3 % (w/v) K-carrageenan gel beads containing B. firmus NRS 783 cells were stored in a 2.1 5% (w/v) KC1 solution at 0 "C. While the percentage of the initialviable cells, present in the cell-carrageenan suspension before gel bead formation, which retained viability was high during the first 3 days, this percentage decreased severely beyond 3 days, with only 1.4% of the initial viable cells being alive after a week of storage of gel beads (see Table I). Retention of cell viability was higher when the gel beads carrying B. firmus NRS 783 cells were stored in the semidefined medium described earlier under Materials and Methods. While a large percentage of viable cells present in the cell-carrageenan suspension prior to formation of gel beads remained viable during the first 5 days, cell viability had declined substantially after 9 days of storage (see Table I). Cell Leakage. The pore size of the gel is a critical parameter in selecting a gel matrix for a particular entrapment process. While a matrix with small pores is attractive due to efficient retention of cells in the gel, it usually limits diffusion of nutrients and products and may thereby reduce the overall productivity. While diffusion of these species in gels with large pores may be easier, significant leakage of cells may occur in such gels (Nguyen and Luong, 1986). Excessive effusion/leakage of cells may cause fracturing of the gels (Wadaet al., 1980),and growth of leaked cells exterior to the support matrix may lead to obstruction of flow within an immobilized cell reactor. Experimental evidence from many previous studies indicates that while the extent of cell leakage from a support matrix increases with cell loading and cell growth rate, cell leakage occurs even a t low and intermediate cell loadings. Cell leakage from gels can be reduced significantly by employing tighter matrices (Cheetham et al., 1979; Nilsson and Ohlson, 1982). The tightness of gels can be controlled by careful selection of gel composition and gel preparation procedure. In the present work, the tightness of the differently prepared gels was compared by examining cell leakage from cell-laden gels and cell penetration into cell-freegels in a nongrowth environment. The motivation for using a nongrowth environment is to study the processes of cell leakage and cell penetration in

50 I

I

PI

s 40

"0

/

I

60 80 100 Time, m i n Figure 5. Concentration profiles of cells leaked from cell-laden gels (gelmorphology: conical discs) when contacted with a saline solution containing 0.8% (w/v) NaCl and 2.1 7% (w/v)KC1. The shake flasks containing cell-laden gels (volume of carrageenan solution in each flask prior to gel formation = 50 mL) and the saline solution (50 mL) were agitated a t 250 rpm and 37 OC on a water bath shaker. The symbols 0,A, 0 , and 0 denote experimental data for 2 % , 3 % , 4 % ,and 5 % (w/v) K-carrageenan gels, respectively. Cell concentration is expressed as 105 viable cells/mL of saline solution.

20

40

the absence of any bias that may be introduced by cell growth and metabolic activities of immobilized cells. Such unbiased investigation is very useful for establishing the relations between (i) cell loading and the extent of cell leakage and (ii) the extent of cell penetration and cell concentration in the gel-free suspension exterior to the gels, as well as for examining the relative importance of cell leakage and cell penetration. Being purely nonreactive processes, the characteristics of cell leakage and cell penetration should be independent of the status of cells. The information obtained on the characteristics of these processes via experiments conducted in a nongrowth environment should therefore be readily applicable to situations involving active cell growth. Leakage of cells from cell-laden gels was studied (i) after curing the gels with a 2.1% (w/v) KC1 solution and (ii) during curing of gels with a 2.1 % (w/v) KC1 solution. In the former situation, the gel-liquid interface in each shake flask was marked before contacting the gel with a saline solutioncontaining0.8% (w/v) NaCland 2.1 % (w/v) KC1. Concentration of viable cells released into the saline solution was measured periodically for up to 90 min after the first contact of the solution with the gel. No detectable change in the position of the gel-liquid interface was observed during this period, indicating that swelling and disintegration of gel, if any, were negligible owing to the presence of KC1 in the saline solution. Concentration profiles of leaked cells are presented in Figure 5 for gels with four different carrageenan contents. The initial concentration of cells in each carrageenan-cell suspension at 45 "C was 1.11X 108viable cells/mL of suspension. Cell leakage is observed to decrease with increasing carrageenan content of gels up to 4 % (w/v). A similar trend was observed in a study involving whole-cell immobilization in calcium alginate gels, with the extent of cell leakage decreasing with alginate content in the range 0.5%-2.5% (w/v) (Nilsson and Ohlson, 1982). The free volume fraction of gels decreases with increasing carrageenan content up to 4% (w/v) (see Figure 2). The extent of cell leakage therefore appears to be related to some extent to the free volume fraction of gels in this carrageenan concentration range. The larger the average pore size of a gel, the easier will be the leakage of cells from the gel and the higher will be the extent of cell leakage. Although measurements of the average pore sizes of the differently

B b t e d " . Rug., 1991, Vol. 7, No. 6

522

prepared gels were not attempted in the present study, it is apparent from the results in Figure 5 that, for the gels prepared from 2%, 3%, and 4% (w/v) K-carrageenan solutions/suspensions, the averagepore size increased with decreasing carrageenan content, and hence, the extent of cell leakage increased in the same order. The extent of cell leakage in 5 % (w/v) carrageenan gels is nearly an order of magnitude greater than the extents of cell leakage in other gels (see Figure 5). While the free volume fraction of 5% (w/v) K-carrageenan gels is significantly greater than that of 4% (w/v) K-carrageenan gels, the difference in the free volume fractions alone cannot account for the tremendous difference in the extents of cell leakage for these (4% and 5%) gels. As discussed earlier, the gelation temperature was fixed at 45 "C in order to maintain high cell viability in the gels. The gel suspension must be kept a t or above a certain minimum temperature (ideal gelation temperature) for formation of a homogeneous gel (gel with uniformly distributed pores). As mentioned earlier, the ideal gelation temperature for gels prepared from cell suspensions containing 5% (w/v) K-carrageenan is greater than 45 "C. Formation of solid gels from 5% (w/v) K-carrageenan solutions/ suspensionsprepared a t 45 "C was rather rapid. Formation of spherical gel beads of desired size from these solutions/ suspensions was therefore difficult due to rapid solidification of these unless injected very rapidly into the curing solution. Further, gel beads formed from these solutions/ suspensions were more difficult to disintegrate than the gel beads formed from solutions/suspensions containing 4 % (w/v) or less carrageenan. In fact, while the free volume fractions of gel beads prepared from solutions containing 2% and 5 % (w/v) marrageenan were nearly identical (see Figure 2), complete disintegration of gel beads formed from 5 % (w/v) carrageenan solution took twice as much time as that required for disintegration of gel beads prepared from 290 (w/v) carrageenan solution. Microscopicobservation of slices of cell-laden gel beads and gel cubes revealed that while the immobilized cells were distributed uniformly in the gels prepared from 276, 350, and 4 % (w/v) K-carrageenan suspensions, the distribution of entrapped cells in the gels prepared from 5% (w/v) K-carrageenansuspensions was nonuniform. Optical micrographs for cell-laden gels prepared from 3%, 4%, and 5 Y (w/v) K-carrageenan suspensions are shown in Figure 6. A significant portion of the cell population in 550 (w/v) carrageenan gels was in the form of cell clusters (clusters of rod-type cells). Aggregation of polymer chain segments and inhomogeneous distribution of cells in 5% (w/v) carrageenan suspensions due to high viscosity of these at 45 "C, a temperature less than the ideal gelation temperature, appear to have led to cell clusters. Such cell clusters were not observed in the other gels. These micrographs suggest that distribution of free space in the 5% (w/v) carrageenan gels is nonuniform. As opposed to gels formed from 270, 370, and 4% (w/v) carrageenan suspensions, where the free space is in the form of smaller pockets/regions distributed uniformly in the gel volume, the free space in gels prepared from 5% (w/v) carrageenan suspensions seems to exist in the form of both smaller and larger pocketslregions spread nonuniformly in the gel volume (see Figure 6). As far as cell retention is concerned, gels with nonuniform distribution of the free space, with a portion of the free space in the form of larger pockets, are less efficient than gels with uniform distribution of porous space in the form of smaller pockets. It is therefore not surprising that the extent of cell leakage in gels prepared from 5 c"c (w/v) carrageenan suspensions

J

f

I

--

~-

Figure 6. Optical micrographs (magnification = 650) of slices of cell-laden gels prepared from cell suspensions containing 3 '6 (top), dCi (middle), and 55. (bottom) (w/v) K-carrageenan. The gels were prepared at 45 "C and cured for 24 h at room temperature (25 "C) in a 2.1% (w/v) KCl solution.

was substantially higher than the extents of cell leakage in other gels. Curing of a K-carrageenan gel with KCl solution over an extended period leads to shrinking of the gel and expulsion of a fraction of water in the porous space, and along with it some cells (in the case of cell-laden gels), from a gel. The concentrations of cells, effused during curing of cell-laden gels (prepared from suspensions containing various concentrations of K-carrageenan) for 24 h, in a 2.1% (w/v) KCl solution are reported in Figure 7. The trend observed here for the dependence of cell effusion during gel curing on carrageenan content of cell-laden gels is identical to that observed earlier in cell leakage experiments (seeFigure 5). The decline in the extent of cell effusion with an increase in K-carrageenan concentration up to 4% (w/v) is associated with decreases in the average pore size, the

523

Biotechnol. Prog., 1991, Vol. 7, No. 0

20

.z 16 E

4

I

2.4

1

A

I

do 2.0

3 m

3 1.6

k

1.2

30 0.8 M I

4 0.4 0.0

Carrageenan content, percent (w/v) Figure 7. Variation in the extentof cell effusion duringextended gel curing (24 h or more) with variation in carrageenan content of the cell-ladengels. The shake flasks containingcell-ladengel beads and the KC1 solution were agitated at 250 rpm and 25 O C on a water bath shaker. The ordinate represents concentration of viable cells effused from 20 mL of gel beads in 200 mL of the 2.1% (w/v) KCl solution (expressed as lo6 cells/mL of KCl solution).

free volume fraction of gels, and the specific external surface area of gel beads that accompany such an increase. The diameter of gel beads was mentioned earlier to increase with increasing carrageenan content of gels. The specific external surface area (external surface area per unit gel volume) of gel beads, which is inversely proportional to bead diameter, therefore is decreased with an increase in carrageenan content. The extent of cell effusion from gel beads or the extent of cell penetration into gel beads is expected to be influenced by the external surface area of gel beads, across which transport of cells occurs. The severalfold increase in cell effusion in a 5 % (w/v) K-carrageenan gel as compared to cell effusion in the other gels is another indicator of the nonhomogeneity (existence of porous space in the form of smaller and larger pockets and its nonuniform distribution in gel volume) of 5% (w/v) K-carrageenan gels. The effect of initial cell loading on cell leakage from KC1-cured 3 % (w/v) K-carrageenan gels during contact with a saline solution containing 0.8% (w/ v) NaCl and 2.1% (w/v) KC1 is illustrated in Figure 8. The extent of cell leakage was observed to increase with increasing initial cell content of the gels. Similarly, the concentration of cells present in a 2.1 % (w/v) KClsolution after extensive curing of 3 % (w/v) carrageenan gel beads increased with an increase in initial cell loading of the gels (cell concentrations in the KC1 solution for initial cell loadings of 0.532 X 108,1.063X lo8, 2.126 X lo8, and 3.189 X 108viablecells/mL of carrageenan-cell suspension prior to gel formation being 0.233 X lo5, 0.213 X lo5, 0.664 X 105,and 1.268 X 105cells/mL, respectively), indicating an increase in the extent of cell effusion during gel curing. On the basis of the results presented in Figure 5, the percentage reductions in cell contents of 2%, 3 % , 4 % , and 570 (w/v) gels owing to cell leakage during the first 90 min of contact with the saline solution were estimated to be approximately 0.155'6, 0.176, 0.045'6, and 1.0376, respectively. I t is evident from this and the discussion earlier that the tightness of the gels decreases (in other words, the average pore size of the gel increases) in the following order of carrageenan content (w/v) of gels: 475, 370, 270, and 5 % . On the basis of the results presented in Figure 8, the percentage reductions (due to cell leakage) in cell contents of 3% (w/v) carrageenan gels with the initial cell loadings of 0.532 X lo8, 1.063 X lo8, 2.126 X lo8, and 3.189 X lo8 viable cells/mL of carrageenan-cell suspension prior to gel formation were estimated to be approximately 0.046 '3, 0.032 % , 0.031 % , and 0.032 % ,

I

I

I

1

I

20

40

60

80

100

Time, min

Figure 8. Concentrationprofiles of cells leaked from cell-laden gels (gel morphology: conical discs) with different initial cell loadings into a saline solution containing0.8% (w/v) NaCl and 2.1% (w/v) KC1. The shake flasks containing cell-laden gels (volume of carrageenan solution in each flask prior to gel formation = 50 mL) and the saline solution (50mL) were agitated at 250 rpm and 37 "C on a water bath shaker. Cell concentration is expressed as 105cells/mL of salinesolution. The carrageenan content of the gels was 3% (w/v). The symbols 0, A, 0,and 0 denote experimentaldata for initial cell loadings of 0.532 X 108, 1.063X 108,2.126 X lo*,and 3.189 X 108 viable cells/mL of carrageenan-cell mixture prior to gel formation, respectively.

respectively, which indicates the near linearity between the extent of cell leakage and the initial cell content of gels. Cell Penetration. Cell-free gel beads prepared from solutions containing 2%, 3%, 4 % , and 5% (w/v) K-carrageenan were transferred to a cell broth (a batch culture of B. firmus near the termination of its growth phase) containing B. firmus at a concentration of 1.84X lo9viable cells/mL. Profiles of concentration of cells penetrated into gel beads are presented in Figure 9. In each case, cell concentration in gel beads increased for 40-50 min, after which the change in cell concentration in gel beads was insignificant (i.e., the gel beads were "saturated" with B. firmus cells). One would expect that the higher the average pore size of a gel, the higher will be the extent of cell penetration into the gel, and at the same time, increasingly difficult will it be for the gel to retain the penetrated cells. As a result, the extent of "net" cell penetration will increase with increasing average pore size up to a critical size and decrease with any further increase in the average pore size owing to increased difficulty encountered by the gel in retaining the penetrated cells. From the results of cell leakage experiments, it was deduced that the average pore size of the gels increases in the following order of carrageenan content (w/v) of the gels: 4%, 3 % , 2%, and 5%. The increases in the net cell penetration rate and the maximum extent of net cell penetration with K-carrageenan content of gels in the order 4 5% , 3% ,and 2 % (w/v) are commensurate with increases in the average pore size of the gels and the specific external surface area of gel beads in the same order. Gels prepared from 5% (w/v) K-carrageenan solutions/suspensions were mentioned earlier to be nonhomogeneous as far as the distribution of free space in gel volume is concerned, with a portion of the free space existing in the form of larger pockets. It is not surprising then that, despite the nearly identical free volume fractions, the extent of cell penetration in 2 76 K-c=rageenan gels is much greater than that in 5% carrageenan gels. A 5 % carrageenan gel is much less effective for cell retention than the other three gels. On the basis of the results presented in Figure 9, the percentage reductions in cell contents of the nongrowing cultures due to penetration of cells into 2 % , 3% , 4% ,and 5 % (w/v) carrageenan gels were estimated to be approx-

524

Biotechnol. hog., 1991, Vol. 7, No. 6

3.0 I

$1 (d

5 2.0

0

1.0

CI CI

/ /

//

al V

0.0

Time, min Figure 9. Profiles of cell contents of gel beads with different carrageenancontentsin cell penetration experiments. The shake flasks containing cell broth (100mL) and initially cell-free gels (40 mL) were agitated at 250 rpm and 37 O C on a water bath shaker. The symbols 0 , A, 0,and + represent experimental data for 2 % , 3 % , 4 % , and 5% (w/v) K-carrageenan gels, respectively. The cell content of the gels is expressed as los viable cells/mL of gel. imately 0.0064 % , 0.0044 5% , 0.00064 % , and 0.00134 % , respectively. The results presented in Figure 9 therefore reinforce the observation made earlier that the tightness of the carrageenan gels increases in the following order of carrageenan content (w/v) of gels: 5 % , 2 % , 3% ,and 4 % . Treatment of gel beads with cross-linking agents such as glutaraldehyde and triethylenetetraamine has been reported to increase gel strength and reduce leakage of cells from cell-laden gels (Chibata et al., 1974; Takata et al., 1977; Chao et al., 1986). In the present study, cell-laden gel beads cured with a 2.1 % (w/v) KC1 solution for 12 h were treated with these cross-linking agents by immersing the gel beads in solutions containing (i) 1%and 2 % (w/v) glutaraldehyde and (ii) 0.02 M and 0.1 M triethylenetetraamine for 10 min. Gel beads treated with any of the four solutions did not contain any viable B. firmus cells. The two cross-linking agents are thus severely toxic to B. firmus cells entrapped in K-carrageenan gels.

Conclusions Since the minimum free volume fraction of gels was in excess of 0.8 for the carrageenan contents investigated, K-carrageenan should be avery suitable support for wholecell immobilization. The gel density increased slightly with increasing carrageenan content. The mechanical strength of the gels increased with increasing carrageenan content and upon extensive curing with a KCl solution. The gel volume decreased upon extensive curing with a KC1 solution. The higher the strength of KCl solution, the greater was the reduction in gel volume. For all carrageenan concentrations investigated, 90-92 % of the cells that were viable prior to immobilization remained viable after formation and extensive curing of gels, illustrating suitability of K-carrageenan as a support matrix for immobilization of B. firmus. Microscopic observation of slices of cell-laden gel particles revealed that the distribution of cells entrapped in the gels prepared from 5 % (w/v) K-carrageenansuspensions was nonuniform, whereas immobilized cells were distributed uniformly in the gels prepared from 2%, 3% and 4% (w/v) K-carrageenan suspensions. Cell leakage from gels during curingand from extensively cured gels during contact with a saline solution and the extent of net cell penetration into cell-free gels decreased in the following order of carrageenan content of gels: 2%, 3 % , and 4%. The extents of cell leakage from 5% (w/v) carrageenan gels, both during gel curing and upon contact

of cured gels with a saline solution, were nearly an order of magnitude higher than the corresponding extents for gels with the other three carrageenan contents, while the extent of net cell penetration into 5% (w/v) cell-free carrageenan gels was considerably low. These substantial differences are attributable to the nonhomogeneity of the 5 % gels as far as the distribution of free space in gel volume is concerned. For comparable contact times (i) between cell-laden gels and a saline solution and (ii) between cellfree gels and a nongrowing culture, cell penetration was observed to be much less significant than cell leakage. Among the gels studied, 4% (w/v) carrageenan gels appear to be the most suitable choice as support matrix for wholecell immobilization owing to attractive features of these such as higher mechanical strength, tighter configuration (reduced cell leakage), and a high free volume fraction (close to 0.8). The information base on physical properties of K-carrageenan gels and cell-gel interactions in these gels generated here will aid in design of support matrices, based on the polysaccharide, for whole-cellimmobilization.

Acknowledgment Financial support received from the Amoco Foundation, Public Health Service (under Public Health Service Biomedical Research Support Grants 2-S07-RR07027-23, 2-S07-RR07027-24, and 2-S07-RR07027-25),and the Galvin Venture Fund of IIT is gratefully acknowledged.

Literature Cited Anderson, N. S.; Campbell, J. W.; Harding, M. M.; Rees, D. A.; Samuel, J. W. B. X-ray diffraction studies of polysaccharide sulphates: double helix models for K- and &-carrageenans.J . Mol. Biol. 1969, 45, 85-99. Arnott, S.; Scott, W. E.; Rees, D. A.; McNab, C. G. A. c-Carrageenan: molecular structure and packing of polysaccharide double helices in oriented fibres of divalent cation salts. J . Mol. Biol. 1974, 90, 253-267. Aunstrup, K.; Farum; Andersen, 0. Enzyme products. U.S. Patent 3,827,933, 1974. Birnbaum, S.; Larsson, P. 0.; Mosbach, K. In Solid Phase Biochemistry: Analytical and Synthetic Aspects; Scouten, W. H., Ed.; Wiley: New York, 1983. Borchard, W. In Chemistry and Technology of Water-soluble Polymers; Finch, C. A., Ed.; Plenum Press: New York, 1983; pp 113-124. Bryce,T. A.; McKinnon, A. A.; Morris, E. R.; Rees, D. A,; Thom, D. Chain conformationsin the sol-gel transitions for polysaccharide systems, and their characterization by spectroscopic methods. Faraday Discuss. Chem. SOC.1974, No. 57, 221229.

Chanzy, H.; Vuong, R. Ultrastructure and morphology of crystalline polysaccharides. In Polysaccharides: Topics in Structure and Morphology; Atkins, E. D. T., Ed.; VCH Publishers, Deerfield Beach, FL, 1985; pp 41-72. Chao, K. C.; Haugen, M. M.; Royer, G. P. Stabilization of K-carrageenan gel with polymeric amines: use of immobilized cells as biocatalysts at elevated temperatures. Biotechnol. Bioeng. 1986,28, 1289-1293. Cheetham,P. S. J.; Blunt, K. W.; Bucke, C. Physical studies on cell immobilizationusing calcium alginate gels. Biotechnol. Bioeng. 1979,21, 2155-2168. Chibata, I.; Tosa, T. Use of immobilized cells. Annu. Rev. Biophys. Bioeng. 1981,10, 197-216. Chibata,I.; Tosa, T.; Sato,T. Immobilized aspartase-containing microbialcells: preparation and enzymaticproperties. Appl. Microbiol. 1974, 27, 878-885. Chibata, I.; Tosa, T.; Fujimura, M. Immobilized living microbial cells. In Annual Reports on Fermentation Processes, 6; Academic Press: New York, 1983; pp 1-22. Dea, I. C. M. Specificityof interactions between polysaccharide helices and @-1,4-linkedpolysaccharides. ACS Symp. Ser. 1981, NO. 150, 439-454.

Biotechnol. hog., 1991, Vol. 7, No. 6

Finch, C. A. In Chemistry and Technology of Water-soluble Polymers; Finch, C. A., Ed.; Plenum Press: New York, 1983; pp 86-89. Harkins, W. D.; Brown, F. E. The determination of surface tension (free surface energy), and weight of falling drops: the surface tension of water and benzene by the capillary height method. J. Am. Chem. SOC.1919,41,499-524. Helmo, K.; Winther-Nielsen, M.; Emborg, C. Protease productivity in chemostat fermentations with retention of biomass on suspended particles. Enzyme Microb. Technol. 1985,7, 443-444. Kalal, J. In Chemistry and Technology of Water-soluble Polymers; Finch, C. A., Ed.; Plenum Press: New York, 1983;pp 71-80. Kalisz, H.M. Microbial proteinases. Adu. Biochem. Eng./Biotechnol. 1988,36, 1-65. Karel, S.F.;Libicki, S. B.; Robertson, C. R. The immobilization Chem. Eng. Sci. 1985, of whole cells: engineering _ principles. _ 40, 1321-1354. Klein, J.;Wagner, F. Methods for the immobilization of microbial cells. Appl. Biochem. Bioeng. 1983,4, 11. Koshcheyenko, K. A,; Turkina, M. V.; Skryabin, G. K. Immobilization of living microbial cells and their application for steroid transformation. Enzyme Microb. Technol. 1983,5, 14-21. McKinnon, A. A.; Rees, D. A.; Williamson, F. B. Coil to double helix transition for a polysaccharide. Chem. Commun. 1969, 701-702. Moon, S.-H.; Parulekar, S. J. A parametric study of protease production in batch and fed-batch cultures of Bacillus firmus. Biotechnol. Bioeng. 1991,37,467-483. Morris, E. R.; Rees, D. A.; Robinson, G. Cation-specific aggregation of carrageenan helices: domain model of polymer gel structure. J. Mol. Biol. 1980,138, 349-362. Nguyen, A-L.; Luong, J. H. T. Diffusion in K-carrageenan gel beads. Biotechnol. Bioeng. 1986,28,1261-1267. Nilsson, I.; Ohlson, S. Columnar denitrification of water by immobilized Pseudomonas denitrificans cells. Eur. J . Appl. Microbiol. Biotechnol, 1982,14,86-90. Norton, I. T.; Goodall, D. M.; Morris, E. R.; Rees, D. A. Dynamics of salt-induced random coil to helix transition in segmented &-carrageenan. Chem. Commun. 1978,515-516. O’Neill, A. N. Derivatives of 4-O-@-~-galactopyranosyl-3,6-anhydro-D-galactosefrom K-carrageenan. J.Am. Chem. SOC.1955, 77,6324-6325. Perry, R. H.; Green, D.; Maloney, J. 0. Perry’s Chemical Engineers’ Handbook, 6th Ed.; McGraw-Hill: New York, 1984; pp 18-48. Priest, F. G. Extracellular enzyme synthesis in the genus Bacillus. Bacteriol. Rev. 1982,21,711-753.

525

Reid, D. S.; Bryce, T. A.; Clark, A. H.; Rees, D. A. Helix-coil transition in gelling polysaccharides. Faraday Discuss. Chem. SOC.1974,No. 57,230-237. Rees, D. A. Shapely polysaccharides. Biochem. J. 1972, 126, 257-273. Rees, D. A. In Carbohydrates;Aspinall, G. O.,Ed.; Butterworths: London, 1975;Vol. 7,pp 251-283. Rees, D. A.; Steele, I. W.; Williamson, F. B. Conformational analysis of polysaccharides. 111. The relation between stereochemistry and properties of some natural polysaccharide sulfates. J. Polymer Sci. C 1969,No. 28,261-276. Rees, D. A.;Morris, E. R.; Thom, D.; Madden, J. K. Shapes and interactions of carbohydrate chains. In The Polysaccharides; Aspinall, G. O., Ed.; Academic Press: New York, 1982;Vol. 1, pp 238-276. Rinaudo, M.; Rochas, C. Investigations on aqueous solution properties of K-carrageenans. ACS Symp. Ser. 1981,No. 150, 367-378. Shiotani, T.; Yamane, T. A horizontal packed-bed bioreactor to reduce COz gas holdup in the continuous production of ethanol by immobilized yeast cells. Eur. J.Appl. Microbiol. Biotechnol. 1981,13, 96-101. Starostina, N. G.; Lusta, K. A,; Fikhte, B. A. Morphological and physiological changes in bacterial cells treated with acrylamide. Eur. J. Appl. Microbiol. Biotechnol. 1983,18,264-270. Takata, I.; Tosa, T.; Chibata, I. Screening of matrix suitable for immobilization of microbial cells. J . Solid Phase Biochem. 1977,2,225-236. Takata, I.; Kayashima, K.; Tosa, T.; Chibata, I. Improvement of stability of fumarase activity by Breuibacterium flauum immobilized with K-carrageenan and polyethyleneimine. J . Ferment. Technol. 1982,60,431-437. Tosa, T.; Sato, T.; Mori, T.; Yamamoto, K.; Takata, I.; Nishida, Y.; Chibata, I. Immobilization of enzymes and microbial cells using carrageenan as matrix. Biotechnol. Bioeng. 1979,21, 1697-1709. Wada, M.; Kato, J.; Chibata, I. Electron microscopic observation on immobilized growing yeast Saccharomyces sp. cells. J . Ferment. Technol. 1980,58,327-332. Wang, H. Y.; Hettwer, D. J. Cell immobilization in K-carrageenan with tricalcium phosphate. Biotechnol. Bioeng. 1982, 24,1827-1897. Woodward, J. In Immobilized Cells and Enzymes. A Practical Approach; Woodward, J., Ed.; IRL Press: Oxford, England, 1985. Accepted September 10,1991. Registry No. .&amageenan, 11114-20-8; potassium chloride, 7447-40-7.