Effects of mixed starter composition on nisin Z production by Lactococcus lactis subsp lactis biovar. diacetylactis UL 719 during production and ripening of Gouda cheese

A starter culture system that produced both acid and nisin at acceptable ra tes in milk for manufacture of Gouda cheese was developed using nisin Z-pro ducing L. lactis subsp. lactis biovar, diacetylactis UL 719 (UL 719) and a commercial Flora Danica (FD) starter culture. Different compositions of mix ed cultures (0, 0.2, 0.4, 0.6 or 0.8% UL 719 with 1.4% FD) were tested for acidification and nisin Z production in milk after 12 h incubation at 30 de grees C. The 0.6/1.4% combination, selected as the optimal mixture of start er cultures, acidified milk to a suitable pH and produced nisin Z at a high concentration of 512 IU/ml. With this optimal combination, FD numbers of c itrate-fermenting and non-fermenting bacteria did not change compared with the control (1.4% FD). However, with 0.8% of L. lactis strain UL 719 and 1. 4% of the FD starter culture, the numbers of citrate-fermenting and non-fer menting bacteria in fermented milk decreased compared with those obtained w hen milk was inoculated with 0.2, 0.4 or 0.6% of UL 719 added to 1.4% FD or control cultures (1.4% FD). Mixed starter culture ratios 0.6/1.4%, 0.4/1.4 % and 0.5/1.4% (UL 719/FD) were used to manufacture nisin Z containing Goud a cheese which was ripened up to 45 weeks. The composition of control chees es made with 1.4% FD, and nisin Z-containing Gouda cheeses were similar wit h respect to percent moisture, fat, salt and protein. During the ripening p eriod, the cell counts observed were approximately two logs higher in chees e made with the 0.6/1.4% mixed starter culture than in control cheese. In e xperimental cheese produced with 0.6/1.4% (UL 719/FD) mixed starter culture , nisin activity increased from 256 IU/g at the end of manufacture to a max imum of 512 IU/g after 6 weeks of ripening; the levels then decreased to 12 8 and 32 IU/g after 27 and 45 weeks of ripening, respectively. In contrast, nisin Z was not detected in experimental cheeses made with 0.4/1.4% or 0.5 /1.4% (UL 719/FD) mixed starters. Using an affinity purified anti-nisin pol yclonal antibody, anti-rabbit gold-conjugate and transmission electron micr oscopy, nisin Z was found to be localized in the cheese matrix, in fat glob ules, in the casein phase and concentrated at the fat-casein interface. Aft er 27 weeks of ripening, nisin Z was detected preferentially in the fat glo bules of the experimental cheese. (C) 2000 Elsevier Science B.V. All rights reserved.