Nineteen strains of lactic acid bacteria were investigated for antioxidative activity. These includedLactobacillus acidophilus B, E, N1, 4356, LA-1, and Farr; Lactobacillus bulgaricus 12 278, 448, 449, Lb, 1006, and 11 842; Streptococcus thermophilus 821, MC, 573, 3641, and 19 987; and Bifidobacterium longum B6 and 15 708. Intracellular cell-free extract of all strains demonstrated antioxidative activity with inhibition rates of ascorbate autoxidation in the range of 7-12%. Antioxidative mechanisms including metal ion chelating ability, scavenge of reactive oxygen species, enzyme inhibition, and reducing activity of intracellular cell-free extract of lactic acid bacteria were studied. S. thermophilus 821 had the highest metal ion chelating ability for Fe(2+), and B. longum 15 708 showed the highest Cu(2+) chelating ability among the 19 strains tested. All strains demonstrated reactive oxygen species scavenging ability. L. acidophilus E showed the highest hydroxyl radical scavenging ability, and B. longum B6 had the best hydrogen peroxide scavenging ability. Reducing activity was also found in all strains. Most of the strains tested demonstrated excellent reducing activity. B. longum B6 showed the highest reducing activity among the 19 strains tested. In enzyme inhibition, superoxide dismutase activity was not found in these 19 strains, and the activity of superoxide dismutase was not induced when metal ion Mn(2+), Fe(2+), or Cu(2+)Zn(2+) was present.

The inhibition of lipid peroxidation by Lactobacillus acidophilus and Bifidobacterium longum was investigated using two lipid model systems. All eight strains, including six strains of L. acidophilus and two strains of B. longum, demonstrated an inhibitory effect on linoleic acid peroxidation. The inhibitory rates on linoleic acid peroxidation ranged from 33 to 46% when 1 mL of intracellular cell-free extract was tested. In the second model system, the cell membrane of osteoblast was used as the source for biological lipid. The results indicated that all strains were able to protect biological lipids from oxidation. The inhibition rates on cell membrane lipid peroxidation ranged from 22 to 37%. The effect of L. acidophilus and B. longum on inhibition of fluorescent tissue pigment accumulation was also obtained for osteoblastic cells. The inhibition rates on fluorescent tissue pigment accumulation ranged from 20 to 39%. The antioxidative effect of each milliliter of intracellular cell-free extract of L. acidophilus and B. longum was equivalent to 104-172 ppm of butylated hydroxytoluene (BHT). These results indicated that all strains demonstrated high antioxidative activity. The scavenging ability of lipid peroxidation products, tert-butyl hydroperoxide and malondialdehyde, was also evaluated. The results showed that L. acidophilus and B. longum were not able to scavenge the tert-butyl hydroperoxide. Nevertheless, malondialdehyde was scavenged well by these strains.

The antioxidative activity of the intracellular extracts of yogurt organisms was investigated. All 11 strains tested, including five strains of Streptococcus thermophilus and six strains of Lactobacillus delbrueckii ssp. bulgaricus, demonstrated an antioxidative effect on the inhibition of linoleic acid peroxidation. The antioxidative effect of intracellular extracts of 10(8) cells of yogurt organisms was equivalent to 25 to 96 ppm butylated hydroxytoluene, which indicated that all strains demonstrated excellent antioxidative activity. The scavenging of reactive oxygen species, hydroxyl radical, and hydrogen peroxide was studied for intracellular extracts of yogurt organisms. All strains showed reactive oxygen species-scavenging ability. Lactobacillus delbrueckii ssp. bulgaricus Lb demonstrated the highest hydroxyl radical-scavenging ability at 234 microM. Streptococcus thermophilus MC and 821 and L. delbrueckii ssp. bulgaricus 448 and 449 scavenged the most hydrogen peroxide at approximately 50 microM. The scavenging ability of lipid peroxidation products, t-butylhydroperoxide and malondialdehyde, was also evaluated. Results showed that the extracts were not able to scavenge the t-butylhydroperoxide. Nevertheless, malondialdehyde was scavenged well by most strains.

The influence of nonfermented milk containing L. acidophilus or L. bulgaricus on lactose utilization by lactose maldigesters was investigated. Nonfermented milks containing L. acidophilus or L. bulgaricus at 10(8) and 10(9) CFU/ml were prepared using 2% low-fat milk. Lactose maldigestion was monitored by measuring breath hydrogen at hourly intervals for 8 hr following consumption of 400 ml of each diet. Nonfermented milk containing L. acidophilus B at 10(8) CFU/ml were not effective in reducing breath hydrogen and symptoms. Nonfermented milk containing L. acidophilus B at 10(9) CFU/ml only slightly decreased breath hydrogen production; however, the symptoms were significantly improved. Nonfermented milks containing L. bulgaricus 449 at 10(8) and 10(9) CFU/ml were effective in reducing breath hydrogen and symptoms. The results for bulgaricus milk were all significant. In this study, L. acidophilus B and L. bulgaricus 449 were chosen because of their similar beta-galactosidase activity and bile sensitivity. L. acidophilus and L. bulgaricus are both thermophilic lactobacilli and an active transport (permease) system is found in both species for lactose transport. The major factor affecting in vivo lactose digestion in this study appears to be the bacterial cell wall/membrane structures. That the cell wall/membrane structures of L. acidophilus are different from those of L. bulgaricus can be indirectly proven by the results of sonication time for maximum beta-galactosidase activity measurement. The results of this study indicate that L. bulgaricus is usually a better choice than L. acidophilus for manufacturing nonfermented milks for lactose maldigesters.