The structure of cheese is an important quality parameter, equal to flavor in the
consumer’s selection of cheese. Cheese structure depends on gross composition including fat, moisture and mineral content as well as grade of proteolysis. As compared to full-fat Cheddar, reduced-fat versions are characterized by a more firm structure, higher risk of bitterness and lower flavor intensity. The bitterness can be reduced by replacing bovine chymosin (BC) in cheese production with camel chymosin (CC), which has a lower general proteolysis. A disadvantage of the lower proteolytic activity of CC could be the need of an extended ripening period to reach a similar cheese structure as in cheeses produced with BC.
The aim of this project was to compensate for the lower proteolytic activity in cheese
produced with CC compared to BC. Selection of dairy lactic acid bacteria (LAB) for
cheese production with high proteolytic activity, especially with the ability to break down the N-terminal of αS1-CN, should contribute to primary casein breakdown.
Differences between reduced-fat Cheddar produced with BC and CC, and
combinations of CC and BC (CC/BC), respectively, were investigated during ripening in
relation to proteolysis, cheese structure and bitterness. The retention of CC in a curd was compared with BC in a model system. LAB were screened for proteolytic activity, with special focus on degradation of the N-terminal part of αS1-CN. Promising LAB cultures were selected and investigated in cheese trials for their ability to influence proteolysis and structure during cheese ripening. In an attempt to improve the screening methods and contribute to the development of a new classification system of Latcococcus lactic strains, the peptide profile formed by selected strains after growth in milk was analyzed and compared to their CEP gene sequences.
Reduced-fat Cheddar cheese produced with CC had lower activity toward the peptide
bond αS1-CN Phe23-Ph24 compared to cheeses with BC and CC/BC. This resulted in lower amounts of peptides derived from αS1-CN (f1-23) by LAB, lower bitterness and harder
cheese structure. No difference in cheese structure was seen between cheeses with
CC/BC compared to CC. Similar differences were observed in the primary casein
breakdown in cheeses produced with CC and BC when and adjunct culture, Lb.
delbrueckii subsp. lactis, was added together with the starter culture. The adjunct culture mediated an increase in the total amount of amino acids as well as a shorter structure. A model system, used to study the retention of chymosin in a curd, showed that the retention of CC was less dependent on pH compared to BC, and the retention of CC was higher than BC in the pH interval 6.0-6.7. This indicated that the lower casein breakdown in cheeses with CC could not be explained by a lower retention of CC in the curd. A 20% lower dosage of CC, however, was used in the cheese production, and this may have had an influence. By screening the ability of LAB (18 Lactococcus lactis subsp. cremoris, 28 L. lactis subsp lactis, 10 thermophilic Lactobacillus strains and 15 frozen direct vat set strains of thermopholic Lactobacillus) to hydrolyse αS1-CN, candidates were selected for cheese-making experiments. None of the selected proteolytic strains contributed significantly to softening the cheese structure. Cheeses with one of the Lb. helveticus strains had a significantly lower level of intact αS1-CN after four weeks of ripening, but not after 17 weeks. The two selected Lb. helveticus strains increased the breakdown of intact β-CN. All cheeses with selected dairy LAB had a higher release of free amino acids compared to the reference cheeses, and this was correlated to a lower strain at fracture and to cheese structure. Three out of five cheeses with selected LAB were perceived as harder than the reference cheeses by the sensory panellists, and two cheeses were not significant in cheese structure as compared to the reference cheeses.
Lc. lactis strains which were previously defined as group d based on their cleavage
specificity towards αS1-CN (f1-23), could be subdivided into three groups. This grouping was seen both in the variation of CEP amino acid sequences, and in the identified peptides after hydrolysis in milk, independently of each other. The subgroups were compared with a three-dimensional computer model of the Lc. lactis CEP using
Streptococcus CEP as a template, and new amino acid positions relevant for the substrate binding region were suggested.
In this PhD project it was not possible to increase the breakdown of intact αS1-CN in
cheese and it was not possible to soften the cheese structure by using selected dairy LAB with the ability to break down αS1-CN in the N-terminal. Cheeses produced with adjunct cultures accelerated the release of free amino acids which reduced the strain at fracture and shortened the cheese structure.