Exopolysaccharide production by Lactobacillus bulgaricus SZ2 was optimized in modified MRS (M 1 -MRS). Maximum EPS production was 74.3 ± 2 mg/L and the optimized values of the three variables predicted for maximum EPS production included a temperature of 38.7 °C, Bacto-casitone and glucose concentrations of 24.5 and 29.6 g/L, respectively. The kinetics of EPS formation and growth were monitored in M 1 -MRS, skimmed milk (SM), skimmed milk plus sucrose (Suc-SM), and skimmed milk containing Bacto-casitone and yeast nitrogen base (BY-SM). EPS production in all the media seemed to be growth-related. The EPS/cell ratios were determined to be 3.12*10 -10 , 1.43*10 -10 , 4.42*10 -11 and 3.16*10 -11 mg/cell, in Suc-SM, SM, M 1 -MRS and BY-SM, respectively, clearly indicating the greater effect of C/N ratio. The effects of fructosyltransferase and sucrose phosphorylase genes inactivation on the sucrose metabolism in L. reuteri LTH5448 strains were compared. Inactivation of ftfA gene had no influence on growth and only a minor impact on sucrose turnover, implying frucotsyltransferase contributes to sucrose utilization and sucrose phosphorylase mediates sucrose metabolism in the absence of frucotsyltransferase activity. In contrast to the wild type and ? scrP strains, ? ftfA mutant did not produced any homoppolysaccharide and fructooligosaccharides, indicating frucotsyltransferase is the only responsible enzyme for EPS and fructooligosacharide formation in L. reuteri LTH5448. Keywords: Exopolysaccharide, L. bulgaricus SZ2, L. reuteri LTH5448, fructosyltransferase and sucrose phosphorylase
