Para-κ-casein during the ripening and storage of low-pH, high-moisture Feta cheese. Issue 2 (22nd May 2018)

• Record Type: Journal Article
• Authors: Alexandraki, Voula; Moatsou, Golfo
• Is Part Of: Journal of dairy research. Volume 85:Issue 2(2018)
• Page Start: 226
• Page End: 231
• Publication Date: 2018-05-22
• Subjects: Para-κ-casein, -- hydrolysis, -- sheep and goat cheese milk, -- Feta; Dairying -- Periodicals; Dairying -- Research -- Periodicals; 636.214205
• DOI: 10.1017/S0022029918000237
• Languages: English
• ISSNs: 0022-0299

Abstract

Abstract : The hypothesis of this research paper was that the physicochemical conditions in a low-pH, high-moisture white brined cheese such as Feta would make para-κ-casein vulnerable to residual chymosin activity during ripening and storage. It was important to address this hypothesis, since cheese para-κ-casein could theoretically be used for the assessment of the origin of cheese milk by means of various analytical methods. Feta cheese was manufactured from sheep milk and from four different mixtures of sheep and goat milk in triplicate. The para-κ-casein of Feta samples taken during 120 d of ripening and storage was estimated by means of cation-exchange HPLC and proteolysis was determined in terms of free amino groups. Despite their similarity, sheep and goat para-κ-casein were efficiently separated and the changes of their chromatographic areas indicated that hydrolysis took place during the first stage of ripening. In accordance to the evolution of free amino groups, para-κ-caseins remained stable thereafter. The hydrolysis pattern was not affected by the composition of the cheese milk mixture and after 120 d at least two thirds of the initial quantity remained intact. Considering the efficient separation of sheep and goat para-κ-caseins and their stability during Feta storage, the same method was used for the evaluation of the percentage of goat milk in the cheese milk. The actual and the estimated percentage of goat milk within the range 0–40 were strongly correlated ( R = 0·997, n = 60) and the standard error of estimation was 0·914.