Volume 62, No. 1, 1997—JOURNAL OF FOOD SCIENCE—171 Exopolysaccharide production from whey lactose by fermentation with Lactobacillus delbrueckii ssp. bulgaricus M. A. GASSEM, K. A. SCHMIDT, and J. F. FRANK ABSTRACT Ropy Lactobacillus delbrueckii ssp. bulgaricus (strain RR) was used for production of exopolysaccharide in sweet whey and simulated whey per- meate (SWP) supplemented with combinations of lactose, KH 2 PO 4 , NH 4 Cl, casamino acids, and mineral salts. Media were incubated at 32, 37, and 44°C for 72h. Periodic adjustment of pH to 6.2 increased viscosity and lactose utilization, and the free galactose and lactic acid in the media. The effect of pH adjustment was greater than that of supple- mentation with nutrients or minerals. Fermentation of supplemented SWP generally produced lower viscosities than did fermentation of sup- plemented sweet whey. After 24h fermentation, viscosity decreased in pH adjusted media. Viscosity of media was highest when incubation was at 32°C and lowest with incubation at 44°C. Key Words: exopolysaccharide, lactobacillus, whey INTRODUCTION MICROBIAL POLYSACCHARIDE could provide a potential source of biopolymers for food and other industries. Cheese whey is produced in high amounts and thus is readily available for use as a substrate for biotechnological processes. Whey has been used for production of ethanol, lactic and citric acids, biomass protein, and food yeast (Kennedy, 1985; Mann, 1987; Moulin and Galzy, 1984). Growing polysaccharide-producing lactic acid bacteria in whey may provide polymers that could be used as food stabilizers. Growth conditions may notably affect produc- tion of exopolysaccharide by mesophilic lactococci (Cerning et al., 1992), Propionibacterium acidipropionici (Racine et al., 1991), and yogurt cultures (Gancel and Novel, 1994; Rasic and Kurmann, 1978). Our objective was to study the conversion of lactose to exopolysaccharide as affected by media supplements, pH, and incubation time and temperature during fermentation with Lactobacillus delbrueckii ssp. bulgaricus RR grown in lac- tose-enriched sweet whey and simulated whey permeate (SWP). MATERIALS & METHODS Bacteria L. delbrueckii ssp. bulgaricus RR (a ropy strain) was provided by Dr. Howard Morris, Dept. of Food Science & Nutrition, Univ. of Minnesota, St. Paul. The culture was stored in 11% (wt/wt) rehydrated NDM at -80°C, and was cultured twice in sterile skim milk after thawing before use in experiments. Media Two primary growth substrates were used. The first was sweet whey, which was obtained from the Univ. of Georgia Creamery from the man- ufacture of Cheddar cheese. The whey was steamed 30 min and then either filtered through cheesecloth or centrifuged at 7000  g for 15 min at 5°C. Lactose (50 g/L) was added to the partially deproteinized whey because preliminary research indicated that this would maximize poly- saccharide production. The whey was divided into 8 aliquots and sup- Author Gassem is with the Dept. of Food Science, King Saud Univ., Saudi Arabia. Author Schmidt is with the Dept. of Animal Science & Industry, Kansas State Univ., Manhattan, KS. Author Frank is with the Dept. of Food Science & Technology, Univ. of Georgia, Athens, GA 30602-2106. Address inquiries to Dr. J. F. Frank. plemented (Table 1). Supplemented media were steamed for 30 min, cooled, and adjusted to pH 6.2 using 2N KOH. Filter-sterilized mineral solution was then added as required. The mineral solution was a mixture of 40.7 mM MgSO 4  7H 2 O, 4.6 mM MnSO 4  7H 2 O, 1.4 mM FeSO 4  7H 2 O, and 0.9 mM CaCl 2  H 2 O prepared as described by Racine et al. (1991). Simulated whey permeate (SWP) media (9 and 10; Table 1) were prepared using the method of Jenness and Koops (1962) with modifi- cation. The SWP was formulated from two solutions: (1) 15.80g of KH 2 PO 4 , 12.00g of K 3 citrate, 21.29g of Na 3 citrate  5H 2 O, 0.80g of K 2 SO 4 , and 6.00g of KCl dissolved in distilled water and made to 200 mL or (2) 8.98g of CaCO 3 and 2.70g of MgCO 3 dissolved in 150 mL of distilled water and 25 mL of concentrated HCl, and then evaporated to dryness at 100°C, dissolved in distilled water, and made to 200 mL. To prepare SWP medium, 20 mL of solution 1 was added to 940 mL distilled water with 0.3g K 2 CO 3 and autoclaved at 121°C for 15 min. Solution 2 was autoclaved separately, and then 20 mL was added. pH was adjusted to 6.5 - 6.6 using 2N KOH. SWP was supplemented with 50 g/L of lactose; media (#9 and 10) were prepared with additional nutrients as described (Table 1). pH was determined using an Orion pH meter model 701A (Orion Research, Cambridge, MA)and monitored during fermentation using a pilot flask. Periodic pH adjustment during fermentation was accom- plished by adding 2N KOH to increase the pH to 6.2 after 4 to 6h incubation. Manual additions of KOH prevented potential losses in vis- cosity which could be associated with continuously stirred mechanical pH-stat fermentors. Fermentation The inoculum was prepared in 11% NDM (wt/vol) incubated at 35°C overnight, and added at 2%. Growth media were prepared in 500-mL flasks and incubated at 32, 37 or 44°C. Fermentations at 44°C were continued for 48h, and those at 32 and 37°C were allowed to continue for 72h. Analysis Samples were immediately analyzed for pH, dry cell weight, and vis- cosity. Samples were frozen (-40°C) for later analysis of lactose, glu- cose, galactose, and lactic acid. Sample pH was measured using an Orion pH meter model 701A (Orion Research, Cambridge, MA). Cell concentration Dry cell weights for SWP and whey media were determined by vac- uum filtering 10 mL of fermentation broth through a previously dried and weighed 0.45 μm porosity cellulosic, white grid filter (MSI, West- boro, MA). Collected cells were washed twice with distilled water (20 mL/wash) to remove residual soluble solids. The filter and cells were then dried under vacuum at 70°C for 24h. Hassan et al. (1996) deter- mined that washing of cells effectively removed associated capsular ma- terial. A standard curve was prepared for rapid estimation of dry cell weight based on absorbance. Cells were harvested after 24h incubation at 32°C in Medium 3 (Table 1) by centrifugation (12,000  g for 15 min), washed twice with 0.85 NaCl (100 mL/wash), and resuspended in 10 mL of 0.85% NaCl. Absorbance at 580 nm was then measured. Viscosity Viscosity was determined as described by Schellhaass and Morris (1985) using a Haake Rotovisco RV20 coaxial cylinder viscometer with NVI sensor system (Haake, Inc., Saddle Brook, NJ) at incubation tem-