Effect of dissolved oxygen on redox potential and milk acidification by lactic acid bacteria isolated from a DL-starter culture

Milk acidification by DL-starter cultures [cultures containing Lactococcus lactis diacetylactis (D) and Leuconostoc (L) species] depends on the oxidation-reduction (redox) potential in milk; however, the mechanisms behind this effect are not completely clear. The objective of this study was to investigate the effect of dissolved oxygen on acidification kinetics and redox potential during milk fermentation by lactic acid bacteria (LAB). Fermentations were conducted by single strains isolated from mixed DL-starter culture, including Lactococcus lactis ssp. lactis, Lactococcus lactis ssp. cremoris, and Leuconostoc mesenteroides ssp. cremoris, by the DL-starter culture, and by the type strains. High and low levels of oxygen were produced by flushing milk with oxygen or nitrogen, respectively. The kinetics of milk acidification was characterized by the maximum rate and time of acidification (Vamax and Tamax), the maximum rate and time of reduction (Vrmax and Trmax), the minimum redox potential (Eh7 final), and time of reaching Eh7 final (Trfinal). Variations in kinetic parameters were observed at both the species and strain levels. Two of the Lc. lactis ssp. lactis strains were not able to lower redox potential to negative values. Kinetic parameters of the DL-starter culture were comparable with the best acidifying and reducing strains, indicating their additive effects. Acidification curves were mostly diauxic at all oxygen levels, displaying 2 maxima of acidification rate: before (aerobic maximum) and after (anaerobic maximum) oxygen depletion. The redox potential decreased concurrently with oxygen consumption and continued to decrease at slower rate until reaching the final values, indicating involvement of both oxygen and microbiological activity in the redox state of milk. Oxygen flushing had a negative effect on reduction and acidification capacity of tested LAB. Reduction was significantly delayed at high initial oxygen, exhibiting longer Trmax, Trfinal, or both. Concurrently, anaerobic acidification rate maximum Vamax was decreased and Tamax was extended. Fermentation kinetics in nitrogen-flushed milk was not statistically different from that in untreated milk except for Lc. lactis ssp. lactis CHCC D2, which showed faster reduction time after nitrogen flushing. This study clarifies the relationship between the redox state in milk and acidification kinetics of the predominant subspecies in DL-starter cultures. This knowledge is important for dairies to ensure optimized, fast, and controlled milk fermentations, leading to greater standardization of dairy products.