HPLC of proteose peptones for evaluating ageing of packaged pasteurized milk

The proteolysis of casein (CN) occurring in packaged pasteurized milk (PM) during refrigerated storage was studied with relation to hygienic and microbiological characteristics of starting raw milk. Six batches of raw milk having standard plate count (SPC) from 1.5×10⁴ to 2.5×10⁵ cfu mL⁻¹ and somatic cell count (SCC) from 1.6×10⁵ to 4.4×10⁵ units mL⁻¹ were pasteurized (73 °C for 15 s), packaged and stored at 4 °C for 12 days. Capillary zone electrophoresis of CN showed breakdown of β-CN in all PM samples during storage. An HPLC method for monitoring proteose peptones (PP) formation was developed. Level of PP in PM samples increased, with keeping time from 667–789 to 947–1383 mg L⁻¹ and PP formation was significantly (P<0.05) related to SCC of starting raw milk. Electrospray ionization–mass spectrometry showed that PP were mainly represented by PP-5 from either A¹ or A² variants of β-CN. Five commercial samples of PM were analysed for PP formation during 14-day storage at 4 °C. Commercial samples prepared by microfiltration process or bactofugation combined with pasteurization showed the slowest formation of PP. The effect of storage temperature on PP formation was evaluated by keeping a conventional PM sample at either 8 or 12 °C for 12 days. Proteolysis of all major CNs upon action of plasmin and bacterial proteinases was observed under these conditions. PP level thus proves to be a reliable analytical index for evaluating the ageing of packaged PM during refrigerated storage.     

Biochemical changes throughout the ripening of a traditional Spanish goat cheese variety (Babia-Laciana)

The physico-chemical characteristics, proteolysis (classical nitrogen fractions, caseins and their degradation products and free amino acids), and lipolysis (fat acidity and free fatty acids) were studied throughout the ripening of three batches of Babia-Laciana cheese, a Spanish traditional variety made from raw goats’ milk. The main compositional characteristics of this cheese at the end of the ripening are its high content of total solids (TS) (78.0±2.4 g 100 g⁻¹ of cheese) and fat (61.1±1.2 g 100 g⁻¹ of TS), the presence of residual lactose (1.6±0.8 g 100 g⁻¹ of TS) and its low content of sodium chloride (1.1±0.7 g 100 g⁻¹ of TS) and ash (2.8±0.5 g 100 g⁻¹ of TS). Its pH values (4.44±0.72) are extraordinarily low. The evolution and final values of the different nitrogen fractions show that this cheese undergoes a very slight proteolysis, a fact which was corroborated when the caseins and their degradation products were quantified: β-casein did not undergo any modification throughout ripening, while only 21% of the α_s-caseins were degraded. Free amino acids content increased by a factor of about 7 throughout ripening, resulting in a high content of γ-amino butyric acid and a low content of glutamic acid at the end of the process. Fat acidity increased very slightly, approximately 4.5 times, during ripening, reaching final values of 3.5±2.2 mg KOH g⁻¹ of fat. The total free fatty acids content showed a similar evolution to fat acidity. At the end of the ripening process, the main free fatty acid was C_18:1, followed by C₁₆ and C₁₀.     

Contribution of proteolytic activity associated with somatic cells in milk to cheese ripening

The effect of proteolytic enzymes from somatic cells on cheese quality was studied. In preliminary experiments, milk and two sodium caseinate systems (pH 6.5 and pH 5.2, the latter in the presence of 5% NaCl) were used as substrates to investigate the proteolytic activity of somatic cells recovered from mastitic milk. Urea-polyacrylamide gel electrophoretograms of hydrolysates suggested that somatic cell extracts contributed directly to proteolysis both in buffer and in milk, but that such activity was reduced by batch pasteurisation (63 °C for 30 min). Sodium caseinate was readily hydrolysed by somatic cell extracts; hydrolysis of α_s1-casein was greater at pH 5.2 and increased with level of somatic cells, suggesting that somatic cells contain proteolytic enzymes which are more active at acidic pH values. Subsequently, miniature Cheddar-type cheeses were made from batches of milk to which somatic cells were added (at levels of levels of 3×10⁵ or 6×10⁵ cells mL⁻¹), either before or after pasteurisation. Proteolysis during ripening of cheese (as measured by levels of pH 4.6-soluble nitrogen) increased with somatic cell addition, although this effect was reduced by pasteurisation after cell addition. Somatic cells may also have directly influenced cheese moisture content, which has been established as a principal indicator of quality of Cheddar-type cheese. Proteolytic enzymes of somatic cells from milk were shown to contribute directly to proteolysis in milk and cheese.     

Effect of commercial adjunct cultures on proteolysis in low-fat Kefalograviera-type cheese

The effect of two commercially available adjunct cultures, LBC 80 (Lactobacillus casei subsp. rhamnosus) and CR-213 (containing Lactococcus lactis subsp. cremoris and Lc. lactis subsp. lactis) on the proteolysis in low-fat hard ewes’ milk cheese of Kefalograviera-type was investigated. Two controls, a full-fat cheese (306 g kg⁻¹ fat, 378 g kg⁻¹ moisture) and a low-fat cheese (97 g kg⁻¹ fat, 486 g kg⁻¹ moisture, made using a modified procedure), were also prepared. The effect of adjunct culture on proteolysis, as examined by polyacrylamide gel electrophoresis of cheese and water soluble cheese extracts, was marginal. The reverse-phase HPLC peptide profiles of the water soluble extracts from low-fat cheeses were similar although some quantitative differences were observed between low-fat control cheese and experimental cheeses. The fat content as reflected by the differences in peptide profiles affected the pattern of proteolysis. Proteolysis, as measured by the percentage of total nitrogen soluble in water or in 120 g L⁻¹ trichloroacetic acid, was significantly (P<0.05) affected by the addition of adjunct cultures. Furthermore, the adjunct cultures enhanced the production of low molecular mass nitrogenous compounds; the levels of total nitrogen, soluble in 50 g L⁻¹ phosphotungstic acid, and of free amino acids were significantly (P<0.05) higher in the low-fat experimental cheeses than in the low-fat control cheese.     

Kinetic parameters of Escherichia coli O157:H7 survival during fermentation of milk and refrigeration of home-made yoghurt

The survival parameters of Escherichia coli O157:H7 during milk fermentation (carried out by the LIM or “longer incubation method” at 30 °C, or by the SIM or “short incubation method” at 43 °C) and storage of home-made yoghurt at refrigeration temperatures (2, 4, or 8 °C) were studied. The E. coli O157:H7 counts increased slightly during fermentation by the LIM, from 5.1 to 5.4 log cfu mL⁻¹, and it was not found after 21 d of storage at 2 or 4 °C, and after 10 d at 8 °C. The microorganism counts increased from 4.8 to 5.4 log cfu mL⁻¹ during the SIM, and it was not detected after 7 d stored at 8 °C. The microorganism grew faster at 43 °C (generation time=0.93 h) than at 30 °C (4.12 h) during the fermentation period. The death time decreased with the increase of the storage temperature (from 38.1 h at 2 °C to 30.1 h at 8 °C) in the yoghurt produced by fermentation at 30 °C; however, a clear relationship between death time and storage temperature was not evident at 43 °C. The pH values of the yoghurt ranged from 4.0 to 4.7.     

Production of traditional Greek yoghurt using Lactobacillus strains with probiotic potential as starter adjuncts

Lactobacillus plantarum ACA-DC 146 and L. paracasei subsp. tolerans ACA-DC 4037 were examined for their potential application as adjuncts in the production of traditional Greek set-type yoghurt. Both strains displayed low milk acidification activity, while no inhibition was observed towards or from the yoghurt starters used. Yoghurt produced with L. paracasei subsp. tolerans ACA-DC 4037 exhibited the best sensory properties, with a rich traditional smooth taste, and the strain was selected for further trials. Yoghurt produced with this strain as an adjunct had good physicochemical properties. After 2 weeks of refrigerated storage, microbial loads (>7.0 log cfu g⁻¹) were in accordance with international recommendations and guidelines for probiotic and starter cultures in milk products. Increasing the microbial load further, using concentrated and encapsulated inocula (10–11 log cfu g⁻¹), gave yoghurt with long fermentation times and poor organoleptic properties.     

Fate of enterotoxigenic Staphylococcus aureus in the presence of nisin-producing Lactococcus lactis strain during manufacture of Jben,a Moroccan traditional fresh cheese

The inhibitory effect of nisin-producing Lactococcus lactis subsp. lactis UL730 on the growth of enterotoxigenic Staphylococcus aureus J10 during manufacture of Jben, a Moroccan traditional fresh cheese prepared from recombined milk, was investigated.With an inoculum level of 10³ cfu mL⁻¹, S. aureus was absent in Jben four days after inoculation when the nisin-producing lactococcus was used as lactic starter. In contrast, it survived after that period, when the starter was non-nisin-producing. No staphylococcal thermonuclease was detected in all Jben samples made from milk inoculated with S. aureus at the level of 10³ cfu mL⁻¹.With a higher inoculum of 10⁵ cfu mL⁻¹, S. aureus was still present in Jben after manufacture and persisted during the storage of the product for 3 days in the laboratory, even when the starter used was nisin-producing. Staphylococcal thermonuclease and type C enterotoxin were detected in all Jben samples made from milk inoculated with 10⁵ cfu mL⁻¹. Thermonuclease and enterotoxin were already produced in the coagulum, at 24 h after milk inoculation with S. aureus.     

Optical instrument for the rapid detection of microorganisms in dairy products

MicroFoss, an optical instrument capable of detecting metabolic changes due to microbial activity, was tested for the detection of various groups of microorganisms in dairy products. Optical changes in the growth medium are monitored in a semi-fluid zone that separates the liquid medium containing the sample. Raw and pasteurised milks were evaluated for total viable counts (TVC), coliform and Enterobacteriaceae. Yoghurt was evaluated for coliform, Enterobacteriaceae and yeast.The MicroFoss TVC method for both pasteurised and raw milk utilised pre-filled vials containing 8 mL of Nutrient Broth with Bromcresol purple with 2 mL of milk sample. In a comparison of the instrument TVC methodology to the standard plate count methodology, correlation coefficients R=−0.93 and R=−0.92 were obtained for raw and pasteurised milk, respectively. An advantage of using the MicroFoss method was that results are obtained in 2–11 h rather than 24–48 h, with samples with bacterial counts above the acceptable levels detecting in most cases within an 8 h shift.The MicroFoss methodology for the estimation of numbers of coliform and Enterobacteriaceae utilises pre-filled vials containing 5 mL of broth. In a comparison between the MicroFoss coliform methodology and the corresponding plate count methodology utilising Violet Red Bile Agar (VRBA), a correlation coefficient of R=−0.95 was obtained. An identical correlation coefficient of R=−0.95 was obtained for the comparison of the MicroFoss Enterobacteriaceae methodology and the corresponding plate count methodology utilising VRBA+1% glucose. In 60 pasteurised milk samples tested, only 2 samples were positive by the MicroFoss method, one of which had 1 cfu mL⁻¹ by the plate method while the other had <1 cfu mL⁻¹. Coliform or Enterobacteriaceae were isolated from the detecting vials demonstrating the higher sensitivity of MicroFoss.The MicroFoss method for the detection of the presence of yeast in yoghurt required a 1:10 dilution of the product followed by a pH adjustment. Presence of yeast in yoghurt and soft cheeses was always detected without any false positive results. Results were available in less than 40 h.     

Polymorphonuclear neutrophils and Escherichia coli proteases involved in proteolysis of casein during experimental E. coli mastitis

An Escherichia coli mastitis model was used to characterize enzymes involved in bovine mammary tissue damage and proteolysis in milk. One-quarter each of four cows were inoculated with a suspension (10⁴ cfu mL⁻¹) of E. coli P4:O32. Blood and milk were collected before inoculation and for 216 h afterwards. Intracellular elastase, collagenase and cathepsin activities were measured by flow cytometry of peripheral blood leukocytes and milk polymorphonuclear neutrophils (PMNs). Leukopenia occurred in peripheral blood 9 h after infection, concomitant with an increase in somatic cell count in milk. Milk PMNs had lower activity of cathepsins and collagenase than peripheral blood PMNs. In parallel, milk samples were studied by zymography, and several proteases were detected in mastitic milk. These activities increased after infection, to reach a peak in 6 h. However, total protease profiles and plasmin activities differed. It was concluded that proteases released by PMNs and E. coli contribute to proteolysis of casein during mastitis, as well as plasmin.     

Evaluation of pulsed electric field and minimal heat treatments for inactivation of pseudomonads and enhancement of milk shelf-life

Pulsed Electric Field (PEF) treatment of milk provides the opportunity to increase the shelf-life of fresh milk for distribution to distant markets. PEF treatments were evaluated in sterile (UHT) milk to determine the inactivation of added spoilage Pseudomonas isolates and the subsequent gains in microbial shelf-life (time taken to reach 10⁷ CFU mL^− 1). Little inactivation of Pseudomonas was achieved at 15 or 40 °C compared with 50 or 55 °C. The greatest inactivation (> 5 logs) was achieved by processing at 55 °C with 31 kV cm^− 1 (139.4 kJ L^− 1). Heat treatment at the application temperature without PEF treatment caused minimal inactivation of Pseudomonas (only 0.2 logs), demonstrating that the inactivation of the Pseudomonas was due to the PEF treatment rather than the heat applied to the milk. At added Pseudomonas levels of 10³ and 10⁵ CFU mL^− 1, the microbial shelf-life of PEF-treated milk was extended by at least 8 days at 4 °C compared with untreated milk. The total microbial shelf-life of the PEF-treated milk was 13 and 11 days for inoculation levels of 10³ and 10⁵ CFU mL^− 1 respectively. The results indicate that PEF treatment is useful for the reduction of pseudomonads, the major spoilage bacteria of milk.Industrial relevancePseudomonads are the major psychrotrophic spoilage microflora of refrigerated, stored HTST pasteurised milk. Long-life (UHT) products are an important component of milk sales in South-East Asia, but in recent years there has been an increasing demand for less processed milk products with extended shelf-life. The recent practice of shipping fresh bulk milk from Australia to South-East Asian countries has necessitated additional heat treatment prior to export and on arrival, to achieve the required shelf-life. Pulsed electric field treatment of HTST milk, applied alone or in combination with mild heat under optimised conditions, offers the opportunity of shelf-life extension, while limiting the reduction in quality attributes of milk associated with more severe additional heat treatments.     

Evolution of lipolysis during the ripening of traditional Feta cheese

Lipolysis was studied during ripening of traditional Feta cheese produced in two small dairies, A and B. The cheeses were made from a thermized mixture of ewes’/goats’ milk by using yoghurt as starter and artisanal rennet from lambs’ and kids’ abomasa (cheese A) or mixed artisanal rennet with calf rennet (cheese B).The acid degree value and the free fatty acids (FFA) contents in both cheeses increased sharply up to 18 d (pre-ripening period at 15 °C) and continued to increase throughout ripening. In both mature cheeses, acetic acid was found at high levels (13–18% of the total FFAs). However, except for this, all FFA contents differed significantly (P < 0.05) between the two cheeses throughout ripening. The levels of individual and total C2:0–C8:0, C10:0–C14:0 and C16:0–C18:2 fatty acids were significantly higher (P < 0.05) in cheese A than in cheese B. Presumably the difference, especially in the C2:0–C8:0 content, was due mainly to the type of the rennet used. Butyric acid was the dominant FFA in cheese A (20% of the total FFAs at 120 d), while the most abundant FFAs in cheese B were capric (18%) and lauric acid (18%). In general, the lipolysis degree of the two cheeses was higher than those reported for the industrially-made Feta cheese.In organoleptic evaluation, cheese A had a piquant taste that was attributed to its high content of butyric acid and showed a significantly (P < 0.05) higher total score than cheese B.     

Immunoglobulins,growth factors and growth hormone in bovine colostrum and the effects of processing

In colostrum collected 0–80 h postpartum the contents of immunoglobulins (Igs), transforming growth factor beta-2 (TGF-β2), insulin-like growth factor-1 (IGF-1) and growth hormone (GH) were analysed. Colostrum initially contained 90 mg mL⁻¹ IgG1, 2.8 mg mL⁻¹ IgG2, 1.6 mg mL⁻¹ IgA, 4.5 mg mL⁻¹ IgM, and these concentrations declined by 92%, 87%, 93% and 84%, respectively, in the samples collected later. Of the growth factors, colostrum initially contained 289–310 ng mL⁻¹ TGF-β2 and the concentration diminished to 66 ng mL⁻¹. The content of IGF-1 and GH postpartum decreased from 870 to 150 ng mL⁻¹, and from 0.17 to <0.03 ng mL⁻¹, respectively. Heat treatment and freeze-drying of colostral whey decreased the content of Igs to 75%, while the contents of IGF-1 and TGF-β2 were unaffected. A similar processing, including filtration steps reduced also the IGF-1 and TGF-β2 by 25%. IgM seems to be the most sensitive of the Igs to processing.     

Purification of angiotensin I-converting enzyme inhibitory peptides and antihypertensive effect of milk produced by protease-facilitated lactic fermentation

Fresh low-fat milk was fermented with five mixed lactic acid bacteria for up to 30 h at 42 °C. A protease, prozyme 6, was added 5 h after the beginning of fermentation. The whey was separated from the fermented milk and freeze-dried. As the fermentation time extended to 30 h, soluble protein content increased from 30.9 to 195.9 mg g⁻¹, free amino acid content increased from 2.8 to 192.8 mg g⁻¹, peptide content increased from 6.4 to 402.8 mg g⁻¹ and γ-aminobutyric acid (GABA) increased from 0 to 80.6 mg 100 g⁻¹, while inhibition of angiotensin I-converting enzyme (ACE) increased as indicated by a decrease of IC₅₀ from 1.18 to 0.24 mg mL⁻¹, respectively. The amino acid sequences of two ACE inhibitory peptides were Gly–Thr–Trp and Gly–Val–Trp, of which the IC₅₀ values were 464.4 and 240.0 μm, respectively. The systolic blood pressure and diastolic blood pressure of spontaneously hypertensive rat (SHR) were reduced 22 and 21.5 mm Hg, respectively, after 8 weeks of oral administration of diluted whey (peptide concentration 5 mg mL⁻¹) from the 30 h fermentation.     

Effects of somatic cell counts on the physicochemical and rheological properties of yoghurt made from sheep’s milk

In the present work, yoghurts were made from sheep’s milk with two different somatic cell count (SCC), at low (200 000 cells mL^?1) and high (750 000 cells mL^?1) levels. The characteristics of the final product were analysed for pH, acidity, protein, total solids, fat, syneresis, water holding capacity (WHC) and apparent viscosity. Samples were analysed on days 1, 7 and 14 after production of yoghurts. The SCC had no significant effect either on the acidity or pH of the yoghurt at 24 h (P > 0.05) but a significant effect (P < 0.05) was observed at 168 h. No effects of SCC were observed on total solids and fat content of the yoghurt after 24 and 168 h. High SCC (HSCC) yoghurt had higher protein content (P < 0.05). The yoghurt with the highest SCC had the highest level of syneresis. Viscosity of HSCC yoghurt was higher than that of the low SCC yoghurt on days 1, 7 and 14 of storage. The flow properties also showed that the low SCC yoghurt was softer than that from milk with high content in somatic cells.     

Holder pasteurization impacts the proteolysis,lipolysis and disintegration of human milk under in vitro dynamic term newborn digestion

When the mother's own human milk is unavailable or limited, pasteurized human milk from milk banks is preferentially administered instead of infant formula, especially for vulnerable hospitalized neonates. Holder pasteurization (62.5 °C, 30 min) may alter human milk composition and structure, which may modulate its digestive behavior. An in vitro dynamic system was set up to simulate the gastrointestinal digestion of term newborns in order to compare the kinetics of lipolysis, proteolysis and structural disintegration of raw versus pasteurized human milk. Human milk from 5 donors was pooled. Half of the pool was either administrated raw (RHM) or pasteurized (PHM). Digestions were conducted at least in duplicate for RHM and PHM. Heat-induced protein aggregation was observed in PHM. During gastric digestion, β-casein was proteolyzed significantly faster for PHM than for RHM (p < 0.05), whereas lactoferrin tended to be proteolyzed slower (p = 0.07) for PHM. Pasteurization selectively affected the intestinal release of some amino acids. At any time of the gastrointestinal digestion, the lipolysis of PHM was significantly lower than that of RHM, but no impact was observed on the profile of released fatty acids. RHM presented a structural destabilization after 60 min of gastric digestion, while there was no large variation for PHM. In the intestinal phase, the evolution of the particle sizes was rather similar. Overall, Holder pasteurization impacted the proteolysis, lipolysis and disintegration of human milk. However, this impact was limited and the physiologic and metabolic consequences remain to be investigated.     

Partial renneting of pasteurised bovine milk: Casein micelle size,heat and storage stability

The effects of partial renneting at low temperature on the casein micelle (CM) size and the storage stability of milk were investigated. Low chymosin concentrations (≤ 0.03 IMCU mL^− 1) was applied to pasteurised skim milk at 4 °C and enzyme activity was terminated by thermal application at 60 °C/3 min and 85 °C/30 min, referred to as low heat (LHT) and high heat (HHT) treatment milk, respectively. The addition of rennet with concentrations of 0.01, 0.02 and 0.03 IMCU mL^− 1 for 15 min resulted in κ-casein hydrolysis of 10, 20 and 25%, respectively. Moreover, mean CM size of milk was reduced by up to 10 nm. For LHT milk, the renneted micelles appeared to be stable for up to 17 days, especially in response to the application of 0.01 IMCU mL^− 1 and at a storage temperature of 4 °C. Severe heating at 85 °C/30 min to inactivate the enzyme caused an increase in CM size.     

Antihypertensive activity of milk fermented by Enterococcus faecalis strains isolated from raw milk

A total of 231 microorganisms were isolated from raw cow milk samples and the angiotensin-converting enzyme-inhibitory (ACEI) activity of the resultant fermented milk produced with the isolated microorganisms was assayed. Forty-six of these microorganisms were selected on the basis of high ACEI activity. Four Enterococcus faecalis strains stood out as producers of fermented milk with potent ACEI activity (IC₅₀ (the protein concentration that inhibits 50% of ACE activity): 34–59 μg mL⁻¹). Single doses (5 mL kg⁻¹) of the whey fraction obtained from these fermented milk samples were administered to spontaneously hypertensive rats (SHR) and to normotensive Wistar-Kyoto (WKY) rats in order to investigate their possible antihypertensive activity. Highly significant decreases in the systolic blood pressure (SBP) and in the diastolic blood pressure (DBP) were observed when the fermented milk was administered to SHR. Nevertheless, the fermented milk did not modify the SBP and the DBP of the WKY rats. Raw cow milk is an excellent source of wild lactic acid bacteria able to produce fermented milk with antihypertensive activity and antihypertensive activity of milk fermented by Enterococcus faecalis strains was associated with peptides different from Ile-Pro-Pro and Val-Pro-Pro.     

Assessing the proteolytic and lipolytic activities of single strains of mesophilic lactobacilli as adjunct cultures using a Caciotta cheese model system

Seventeen strains of mesophilic lactic acid bacteria, isolated from cheese (non-starter lactic acid bacteria, NSLAB) or sourdough, were used individually as adjunct cultures in a Caciotta cheese model system. Adjunct cultures were monitored by randomly amplified polymorphic DNA analysis and their cell counts mainly varied from ca. 9.0 to 8.0 log cfu g⁻¹ throughout 36 days of ripening. Adjunct cultures influenced differently cheese proteolysis. Both NSLAB and sourdough strains caused an extensive secondary proteolysis; however, some NSLAB strains produced the highest concentration of free amino acids. Principal component analysis (PCA) differentiated cheeses manufactured with NSLAB strains Lactobacillus parabuckneri B₉F_ST, Lb. paracasei B₆₁F₅, Lb. curvatus 2768 and Lb. rhamnosus ATCC 7469 based on the accumulation of Lys, Glu, Phe, Hist, Asp and Met. Assessment of cheese lipolysis showed that: (i) highest concentrations of free fatty acids (FFA) were found with NSLAB strains Lb. rhamnosus ATCC 7469 and Lb. casei subsp. pseudoplantarum 2742 (ca. 10 500 mg kg⁻¹); (ii) PCA differentiated cheeses manufactured with NSLAB strains Lb. rhamnosus ATCC 7469 and Lb. casei subsp. pseudoplantarum 2742 based on the accumulation of palmitic (C16:0) and linoleic (C18:2) acids, and those with Lb. curvatus 2768 and Lb. parabuckneri B₉F_ST based on the high concentration of short chain FFA; (iii) the cheese made with sourdough strain Lb. sanfranciscensis CB1 had the highest levels of unsaturated FFA.     

Characterization of bacteriocin ST8KF produced by a kefir isolate Lactobacillus plantarum ST8KF

Lactobacillus plantarum ST8KF, isolated from kefir, produced a 3.5 kDa bacteriocin (bacST8KF) active against Lb. casei, Lb. salivarius, Lb. curvatus and Listeria innocua. BacST8KF was sensitive to proteolytic enzymes, but stable between pH 2.0 and 10.0, and heat resistant (20 min at 121 °C). BacST8KF did not adsorb to the surface of the producer cell. Maximum activity (25,600 AU mL⁻¹) was recorded in MRS broth with glucose, in MRS broth with glucose replaced by sucrose, and in MRS broth with glucose, supplemented with KH₂PO₄ after 24 h at 30 °C. Tri-ammonium citrate and glycerol in excess of 5.0 g L⁻¹ repressed bacST8KF production. Production of bacST8KF increased from 800 AU mL⁻¹ after 3 h of fermentation in MRS broth at 30 °C to 12,800 AU mL⁻¹ after 9 h and to 51,200 AU mL⁻¹ after 27 h. These results suggest that bacST8KF may be a secondary metabolite and shows that its mode of activity is bacteriostatic.