Plasmin levels in fresh milk whey and commercial whey protein products

Growth of psychrotrophic bacteria in nonfat dry milk at refrigeration temperatures was shown previously in our laboratory to cause a shift in plasmin (a native milk protease) from the casein to the whey fraction. The whey fraction from cheesemaking is commonly used to make whey protein concentrates and isolates, which then are used as functional ingredients in various food systems. Plasmin activity in whey protein products may cause breakdown of food proteins to have desirable or undesirable effects on food quality. This raised questions about the level of plasmin in commercial whey protein products and factors that affect this plasmin level. Therefore, the objectives of this study were to determine: 1) plasmin concentrations in sweet and acid whey protein products as influenced by Pseudomonas growth during storage of fresh milk, and 2) plasmin concentrations in commercial whey protein products. Whey type (sweet or acid) had a significantly (P < 0.05) greater effect on whey-associated plasmin activity than did Pseudomonas fluorescens M 3/6 growth. Acid whey protein products had significantly (P < 0.05) higher plasmin concentrations than sweet whey. Plasmin activities associated with acid and sweet whey protein products were both significantly (P < 0.0001) affected by the growth of Pseudomonas fluorescens M 3/6. The interaction effect between bacterial growth and whey type on plasmin activity was not significant (P = 0.2457). Plasmin activity in the reconstituted commercial whey protein concentrates (i.e., sweet and acid) varied considerably (16.3 to 330 micrograms/g of protein), but was significantly lower (2.1 to 4.4 micrograms/g of protein, P < 0.05) in whey isolates. These quantitative data were supported by plasmin activity visualized by casein SDS-PAGE.     

Production of proteases by psychrotrophic microorganisms.

Six milk-derived psychrotrophic microbial cultures were screened for the ability to grow at refrigerated temperatures and produce proteases in reconstituted skim milk. Of these, two cultures, Pseudomonas fluorescens M3/6 and Pseudomonas fragi K122, produced extracellular protease(s) beginning 7 d postinoculation when the cultures had entered late log or early stationary phases of growth. Further work with these two cultures showed that intracellular proteases were present after only 20-h incubation, before detection of the extracellular proteases. Using H-D-valyl-L-leucyl-L-lysyl-4-nitroanilide (S-2251), a sensitive substrate for plasmin activity, P. fluorescens was shown to have greater intracellular proteolytic activity than extracellular activity at 20 h of incubation. The intracellular enzyme activity remained constant while the extracellular and periplasmic activities increased over the remaining 6-d incubation period. The proteases in crude extracellular extracts from both cultures were characterized and were heat stable with broad temperature (7 to 52 degrees C) and pH (pH 5.5 to 8.5) ranges for activity and were inhibited by the metal chelator, EDTA, indicating that they were metalloproteases.     

Assay of proteinases of psychrotrophic bacteria in milk using hide powder azure and a simple procedure for clarification of milk

The clarification of milk by addition of solutions of Triton X-100 and EDTA after digestion of added Hide Powder Azure (HPA) was found to provide a simple method of determining the extracellular proteinase activity of Gram-negative psychrotrophic bacteria in pasteurized whole milk. The light absorbance of the clarified HPA digestion product was measured directly, after a brief incubation period, and was stable to storage of samples in diffuse daylight for at least 2 d. Proteinase produced by growth in refrigerated whole milk of as few as 2.5 X 10(6) cfu ml-1 of Pseudomonas fluorescens AR11 was detected.     

Effects of Pseudomonas fluorescens M3/6 bacterial protease on plasmin system and plasminogen activation

Heat-stable proteases produced by the psychrotroph Pseudomonas fluorescens M3/6 have been shown to affect the plasmin system in milk, which in turn will affect the quality of processed milk. The M3/6 proteases cause dissociation of plasmin from casein in minimally processed milk. The objective of this work was to study the effect of M3/6 protease on the plasmin system, as well as its role in plasminogen activation, under commonly applied cheese-making conditions. Isolated M3/6 protease was added to raw milk, which then was pasteurized, and subjected to pH adjustments and CaCl₂ addition. Casein and whey fractions were separated by chymosin treatment then analyzed for plasmin activity. Individual and interaction effects of M3/6 protease addition, pH treatment, and CaCl₂ addition on plasmin activity were studied. Enzyme activity assays were carried out to study individually the effect of M3/6 protease on plasmin system components. Kinetic parameters were calculated to characterize the effect of M3/6 protease on plasminogen activation. Plasmin activity increased in the curd fractions of the protease-treated milk that was subjected to conditions most resembling cheese-making conditions, indicating that M3/6 protease triggered plasminogen activation rather than dissociation of plasmin from casein micelles. Results from the studies on plasminogen activation confirmed that the observed activation of plasminogen in protease-treated samples subjected to cheese making conditions was attributed to the stimulatory effect M3/6 protease had on plasminogen activators (PA). The M3/6 protease stimulated human and bovine PA by increasing their activity 4.5- and 2.5-fold, respectively. Similarly, the catalytic efficiencies of human urokinase-type PA and bovine PA were increased in the presence of M3/6 protease by 12- and 4-fold, respectively. Our research presented a basic step toward fully understanding the effect of bacterial proteases under different processing conditions, where the gathered information can aid in better control of processing conditions based on the desired outcome.     

Volatile organic compounds associated with milk spoilage by psychrotrophic bacteria

The volatile organic compounds of milk contaminated with psychrotrophic bacteria were studied by HS‐SPME and GC/MS. Pseudomonas fluorescens PS14, Pseudomonas fragi PS55, Pseudomonas mosselii PS39, Pseudomonas rhodesiae PS62 and Serratia marcescens S92 were inoculated in sterilised milk (2.5% fat) stored at either 5 °C or 10 °C. A total of 47 volatile organic compounds (VOCs) belonging to seven chemical groups were identified in the spoiled milk. Volatile organic compound patterns peculiar to the inoculate bacterial strains were highlighted. 3‐Methylbutan‐1‐ol, 2 methylpropan‐1‐ol, 3‐hydroxybutan‐2‐one, butan‐2,3‐dione, butanoic and hexanoic acids were revealed as potential chemical spoilage indexes of milk spoilage due to the activity of the five psychrotrophic strains studied.     

Studies of plasmin activity in whey

The presence of the indigenous milk alkaline proteinase, plasmin, in whey products may adversely affect the quality of food products that utilise such ingredients; factors promoting the release of plasmin into whey were therefore investigated. Acidification of pasteurised skim milk (PSM) with glucono-δ-lactone (GDL) or HCl resulted in acid whey with significantly higher (P<0.05) plasmin activity than PSM. Plasmin activity in such whey was inversely correlated with rate of acidification; thus, activity in the whey made with GDL was significantly higher (P<0.05) than in whey made with HCl. Negligible levels of plasminogen in acid whey suggested that activation of the zymogen, plasminogen, contributed to the elevated plasmin activity observed. Sweet whey, manufactured by addition of rennet to PSM, had plasmin and plasminogen-derived activities significantly lower (P<0.001) than those in PSM; the release of plasmin from renneted casein micelles into whey increased in a linear manner with cooking temperature in the range 45–65°C. However, dissociation of plasmin from casein micelles into acid whey reached a maximum at a cooking temperature of 50°C. Analysis of plasmin and plasminogen levels in ultracentrifugal supernatants of pH-adjusted milk indicated that plasmin was released from the micelles at pH>5.0, while plasminogen was released more gradually over the pH range 6.6–4.6. Overall, the technology of manufacture has significant implications for plasmin activity in whey and hence, inversely, for casein products and cheese.     

Milk plasmin, N-acetyl-beta-D-glucosaminidase, and antitrypsin as determinants of bacterial replication rates in whey

Quarter milk samples were collected monthly on a selected herd of 80 Ayrshire cows having a high frequency of subclinical mastitis. Analysis of bacterial growth rates in milk showed that whey prepared from infected or inflamed quarters stimulated bacterial growth. Milk N-acetyl-beta-D-glucosaminidase, antitrypsin, and plasmin activities all showed positive correlations with bacterial replication rates (Staphylococcus aureus and Escherichia coli) in respective whey samples as determined by a turbidometric micro-technique. Increased bacterial replication rates in mastitic whey represent an increased yield of the key nutrients for bacteria. Bacterial growth enhancement can be partly explained by proteose-peptone originating from plasmin activation and casein degradation. However, as multiple regression analysis showed that a combination of the predictor variables: antitrypsin, N-acetyl-beta-D-glucosaminidase and plasmin explained enhancement of bacterial growth better than plasmin alone, other factors connected with inflammation should be sought when searching for growth-enhancing factors in whey. Milk plasmin activities showed increasing activities toward end of lactation (before drying off) as well as during later lactation (age of cow in years minus 2). Bacterial replication was enhanced in parallel with these changes in plasmin activities.     

Effects of somatic cell count and stage of lactation on the plasmin activity and cheese-making properties of ewe milk

The experiment was conducted from March to July 2002 using 5 intensively managed flocks of Southern Italy. In each flock, 2 groups of 50 ewes were created. The groups were designated LSCC (low somatic cell count [SCC]) when their milk SCC was lower than 500,000/mL and HSCC (high SCC) when their milk SCC was higher than 1,000,000/mL. Bulk milk and whey samples were analyzed for fat, total protein, lactose, casein, and whey protein contents. Renneting properties of milk were also determined. Moisture, NaCl, and nitrogen fractions were determined in fresh cheese curds. In addition, plasmin (PL) and plasminogen (PG) activities in milk and cheese were monitored. The proteolytic activity of plasmin by urea-polyacrylamide gel electrophoresis and the white blood cell (WBC) differentials were determined. The HSCC resulted in higher pH values in milk and in higher moisture and lower fat contents in fresh cheese curds. Moreover, a lower recovery of fat and whey proteins was obtained from the HSCC than from the LSCC raw milk. The crude protein and casein contents were higher in the HSCC than in the LSCC curds during early and midlactation; an opposite trend was observed in late lactation. Plasmin and PG activities underwent more marked fluctuations in the LSCC than in the HSCC curds through lactation. The results of this experiment demonstrate that the PL activity in ewe milk is markedly influenced by the SCC, although SCC is not the only parameter for predicting PL and PG evolution in ewe milk. The LSCC milk resulted in a higher proteolytic potential of Canestrato pugliese cheese curds.     

Proteolysis in milk: the significance of proteinases originating from milk leucocytes and a comparison of the action of leucocyte, bacterial and natural milk proteinases on casein

The caseinolytic activities at pH 6.8 of polymorphonuclear (PMN) and mononuclear leucocyte homogenates (equivalent to a level of 10(6) cells/ml milk) were less than the levels of natural milk proteinase activity found in milk from healthy cows. Bulk milks contained approximately 4 times more milk proteinase activity than the composite milks from individual healthy cows. Isolated blood leucocytes, when added to raw milk of good bacteriological quality and stored at 5 degrees C, did not readily degenerate and had no detectable effect on the milk proteins even when these cells were completely disrupted by homogenization of the milk. Pasteurization of milk which contained leucocytes caused loss of cell vitality. Extracellular proteinases of psychrotrophic bacteria growing in milk were not detected until the early stationary phase of growth. The total viable count at which this occurred varied greatly. Proteinase production by a pure culture of Pseudomonas fluorescens was not detected in milk stored at 5 degrees C until a viable count of approximately 10(9) colony forming units (c.f.u.)/ml was obtained, whilst normal bulk milks stored at 5 degrees C produced detectable levels of extracellular proteinase(s) when the psychrotrophic flora reached 10(7)-10(8) c.f.u./ml. Casein proteolysis by PMN and mononuclear leucocyte homogenates resulted in similar polypeptide maps, but plasmin and bacterial proteinase isolated from a strain of Serratia marcescens resulted in polypeptide maps different from each other and from that produced by the leucocyte proteinase(s). The rate of proteolysis of caseins by the different proteinase sources appeared to be in the order alpha s1- greater than beta- greater than greater than kappa-casein for the leucocyte extracts, beta- greater than alpha s1- greater than greater than greater than kappa-casein for bovine plasmin and beta- approximately kappa- greater than alpha s1-casein for for S. marcescens proteinase.     

Differential behavior of Lactobacillus helveticus B1929 and ATCC 15009 on the hydrolysis and angiotensin-I-converting enzyme inhibition activity of fermented ultra-high-temperature milk and nonfat dried milk powder

Consumers' growing interest in fermented dairy foods necessitates research on a wide array of lactic acid bacterial strains to be explored and used. This study aimed to investigate the differences in the proteolytic capacity of Lactobacillus helveticus strains B1929 and ATCC 15009 on the fermentation of commercial ultra-pasteurized (UHT) skim milk and reconstituted nonfat dried milk powder (at a comparable protein concentration, 4%). The antihypertensive properties of the fermented milk, measured by angiotensin-I-converting enzyme inhibitory (ACE-I) activity, were compared. The B1929 strain lowered the pH of the milk to 4.13 ± 0.09 at 37°C after 24 h, whereas ATCC 15009 needed 48 h to drop the pH to 4.70 ± 0.18 at 37°C. Two soluble protein fractions, one (CFS1) obtained after fermentation (acidic conditions) and the other (CFS2) after the neutralization (pH 6.70) of the pellet from CFS1 separation, were analyzed for d-/l-lactic acid production, protein concentration, the degree of protein hydrolysis, and ACE-I activity. The CFS1 fractions, dominated by whey proteins, demonstrated a greater degree of protein hydrolysis (7.9%) than CFS2. On the other hand, CFS2, mainly casein proteins, showed a higher level of ACE-I activity (33.8%) than CFS1. Significant differences were also found in the d- and l-lactic acid produced by the UHT milk between the 2 strains. These results attest that milk casein proteins possessed more detectable ACE-I activity than whey fractions, even without a measurable degree of hydrolysis. Findings from this study suggest that careful consideration must be given when selecting the bacterial strain and milk substrate for fermentation.     

Effect of quorum sensing agents on the growth kinetics of Pseudomonas spp. of raw milk origin

Psychrotrophs, particularly strains of Pseudomonas fluorescens, dominate the microflora of refrigerated raw milk and secrete heat-stable extracellular enzymes (proteases and lipases) which survive pasteurisation and even UHT heat treatments and degrade the casein and fat components of raw milk causing a reduction in cheese yield, gelation of UHT milk and off flavours in many dairy products. These enzymes are usually produced in the late log/early stationary growth phases when the cell density is high. This fact indicated that induction of these enzymes may be a candidate for quorum sensing, a phenomenon by which bacteria can sense and respond to cell population size by means of chemical signals based on the homoserine lactone molecule. It is assumed that the production of these enzymes contributes to the selective advantage of psychrotrophs and hence have a significant effect on their growth kinetics. In the work reported here nine proprietary homoserine lactone compounds were screened, using water as a control, for their effect on the lag phase duration (LPD) and exponential growth rate (EGR) of three strains of Pseudomonas fluorescens isolated from refrigerated raw milk after 1, 3 and 5 days storage. Two compounds viz. N-benzoyloxycarbonyl-L-homoserine lactone and N-3-oxyhexanoyl-DL-homoserine lactone were found to significantly (P < 0.001) reduce the LPD and increase the EGR of the three strains. Further work with these compounds is warranted, monitoring growth in parallel with enzyme production, to determine the extent to which extracellular enzyme production is mediated by the quorum sensing phenomenon in this species.     

Study of the microbial ecology of wild and aquacultured Tunisian fresh fish

Eighty samples of fresh fish were collected in Tunisia and analyzed for microbial load. Quality and hygienic safety of the meat and intestines of wild and aquacultured fresh fish were determined. The mesophilic aerobic plate count and populations of psychrotrophic lactic acid bacteria (LAB) and other psychrotrophic bacteria ranged from 5.67 to 7.29, 4.51 to 6, and 5.07 to 6.21 log CFU/g, respectively. For all microbiological determinations, bacterial counts were lower in meat than in the intestines of fresh fish. For all samples lower microbial populations were found in most of the wild fish than in the aquacultured fish. No isolates of the pathogenic genera Salmonella and Listeria were detected in any sample. Among the 160 strains of biopreservative psychrotrophic LAB and the 150 strains of spoilage psychrotrophic gram-negative bacteria identified by biochemical and molecular methods, Lactobacillus (six species) and Pseudomonas (six species) predominated. Lactococcus, Leuconostoc, Carnobacterium (C. piscicola and C. divergens), Aeromonas, and Photobacterium were the most common genera, and Lactococcus lactis, Lactobacillus plantarum, Pseudomonas fluorescens, and Aeromonas hydrophila were the most common species. These findings indicate that the microbiological quality of fresh fish in Tunisia can be preserved by controlling pathogenic and psychrotrophic bacteria.     

Polyclonal antibodies against protein F from the cell envelope of Pseudomonas fluorescens for the detection of psychrotrophic bacteria in refrigerated meat using an indirect ELISA

An indirect enzyme-linked immunosorbent assay (ELISA) has been developed for detection of Pseudomonas fluorescens and related psychrotrophic bacteria in refrigerated meat. The ELISA uses polyclonal antibodies raised in rabbits against protein F from the cell envelope of Pseudomonas fluorescens AH-70. The anti-protein F antibodies were recovered from the crude antiserum by ammonium sulfate precipitation and conjugated to biotin. Commercial Extr Avidin-peroxidase conjugate was used to detect the biotinylated antibodies bound to their specific antigens. Subsequent enzymatic conversion of substrate gave distinct absorbance differences when assaying meat samples containing P. fluorescens strains of different origin as well as related psychrotrophic microorganisms. The detection threshold for the ELISA assay developed in this work was 10(4)-10(5) cfu cm(-2).     

The effect of refrigerated storage of raw milk on the physicochemical and microbiological quality of Tunisian semihard Gouda‐type cheese during ripening

A Tunisian semihard Gouda‐type cheese made from milk kept at 4 °C for 24, 48, 72 and 96 h was monitored during 45 days of ripening. The effect of milk refrigeration on the evolution of physicochemical parameters in relation to the quantitative variation of the microbial population during ripening of Gouda‐type cheese was investigated. Microbiological and physicochemical analyses were performed on raw milk and cheese samples after curding, 2, 9, 16, 23, 30, 37 and 45 days of ripening time. The raw milk kept under refrigeration at 4 °C for 96 h showed the highest microbial count and proteolysis level. The duration of storage significantly reduced the cheese yield as a result of important solubilisation casein in proteoses‐peptones. Results of different nitrogenous fractions by Kjeldahl method showed enzymatic hydrolysis products of casein whose intensity depended on the maturing stage as well as the refrigeration time. Besides the evident action of the plasmin, original milk protease, on the hydrolysis of casein in soluble fractions, the proteolysis of cheese caseins is also initiated by proteolytic action of the chymosin and extracellular heat‐resistant proteases notably produced by the same psychrotrophic microflora. Lactic acid bacteria starters that constitute the dominant microflora of this type of cheese are also considered as aroma precursors.     

Potential applications of high pressure homogenisation in processing of liquid milk

Studies of the potential of high pressure homogenisation (HPH) for the combined pasteurisation/ homogenisation of raw bovine milk were undertaken. Raw milk was preheated to 45 degrees C and HPH-treated at 150, 200 or 250 MPa; milk outlet temperature at these pressures were 67, 76.8 and 83.6 degrees C, respectively, with a holding time of approximately 20 s. Raw and commercially pasteurized and homogenized (CPH) milk samples were analysed as controls. Fat globules in HPH samples were approximately half the size of those in CPH samples, although differences were not significant (P>0.05). beta-Lactoglobulin was denatured at pressures > or =150MPa, although little denaturation of alpha-lactalbumin was observed. Numbers of psychrotrophic bacteria in raw milk were reduced by 2.73 log cycles by HPH at 150 MPa and were uncountable following HPH at 200 or 250 MPa. Mesophilic bacterial counts were reduced by 1.30, 1.83 and 3.06 log cycles by HPH at 150, 200 or 250 MPa, respectively. No viable Staphylococcus aureus nor coliform cells remained in any HPH milk samples. HPH did not affect the colour of milk and HPH samples did not cream during refrigerated storage. The activities of plasmin, alkaline phosphatase and lactoperoxidase in milk were all greatly reduced by HPH. Pseudomonas fluorescens, inoculated into milk (approximately 10(6) cfu/ml), was reduced to undetectable levels by HPH at 200MPa (milk inlet temperature, approximately 10 degrees C); however, Ps. fluorescens proteinase was quite resistant to HPH under such conditions. Overall, owing to the significant increase in temperature and the possibility of varying the holding time, there may be potential applications for HPH as a novel liquid milk processing technique, combining many advantages of conventional homogenization and pasteurization of milk in a single process.     

Invited review: Plasmin protease in milk: current knowledge and relevance to dairy industry

Plasmin is by far the predominant and most completely studied endogenous protease in bovine milk. Plasmin-induced proteolysis can have either beneficial or detrimental effects on the texture and flavor of dairy products, depending on the extent of hydrolysis and type of dairy product. In cheese, the breakdown of protein can help develop desirable flavors and texture during ripening, whereas in pasteurized milk and ultra-high-temperature milk, proteolysis causes undesirable gelation. Plasmin is part of a complex protease-protease inhibitor system in milk that consists of active and inactive forms of the enzyme, activators, and inhibitors. Considerable research has been done to isolate and characterize components of the plasmin system, determine how they interact, develop and compare quantitation methods, and determine how they are affected by cow characteristics, processing conditions, other milk components, storage conditions, and bacterial proteases. Considerable research has focused on enhancing or minimizing the activity of plasmin system components. The intent has been to control protease activity in casein and whey fractions, depending on the final food or ingredient application. Controlling the activity of plasmin has a great potential to improve dairy product quality and reduce their processing costs.     

Effect of skimmed milk and its fractions on the inactivation of Escherichia coli K12 by high hydrostatic pressure treatment

We investigated the effects of skimmed milk and its protein fractions (casein, whey, globulin, and albumin) on the injury and inactivation of Escherichia coli K-12 by high hydrostatic pressure (HHP) treatment. The protective effect of skimmed milk on HHP-mediated inactivation and injury of E. coli increased with increases in the skimmed milk concentration. However, protein fractions derived from skimmed milk did not exhibit this protective effect. Microscopy analysis by DAPI/PI staining indicated that some cells were localized in the solid portion of skimmed milk, and some of these cells were alive. The coagulated fraction derived from the autoclaved whey fraction also showed a significant protective effect. We speculate that the solid portion in skimmed milk could provide the protective effect to bacterial cells.     

Localization of Mycobacterium avium subspecies paratuberculosis in artificially inoculated milk and colostrum by fractionation

The purpose of the study was to determine the distribution of Mycobacterium avium ssp. paratuberculosis (MAP) across the main milk and colostrum fractions (cream, curd, and whey). Raw milk and colostrum were inoculated with 1 of 2 MAP strains, ATCC 19698 or S-23, yielding initial concentrations of 10⁶ to 10⁷ cfu/mL. After fractionation, for milk as well as for colostrum, 80 to 90% of the recovered MAP cells were found in the curd fraction and 10 to 20% in the cream fraction. Total MAP colony counts in milk whey were 4 to 5 log₁₀ units lower than colony counts of inoculated milk. In colostrum, colony counts were 2 to 3 log₁₀ units lower in whey than in inoculated colostrum. Because of the slow growth of MAP and to proceed more smoothly with set-up and optimization of the method, luminescent MAP strains were used. The high correlation coefficient (r = 0.960) between colony counts and luminescence measurements showed that the use of luminescent MAP strains during method development was plausible.     

GROWTH OF ESCHERICHIA COLI O157:H7 IN THE PRESENCE OF PSEUDOMONAS FLUORESCENS IN SKIMMED MILK AT 7 OR 25C

Escherichia coli O157:H7 STCC 4076 , E. coli O157:H7 STCC 4267 and E. coli STCC 515 were cultured alone or in combination with Pseudomonas fluorescens STCC 378 at 7C or 25C in 10% reconstituted skimmed milk. Culture pH and bacterial population densities were monitored over 40 days. Both E. coli O157:H7 strains grew well after 40 days of incubation at 7C with final pH values between 4.86–4.53. At 25C, both E. coli O157:H7 strains grew during 20 days with final pH values of 4.00–4.14. The pH of the different cultures of this study decreased more at 25C than at 7C. The results suggest that P. fluorescens may inhibit the growth of the other bacteria present in milk at 7C, but this inhibition is weak. In contrast, the growth of E. coli O157:H7 strains in the presence of P. fluorescens appears to be slightly enhanced during most of the incubation period at 25C.     

Activation of transforming growth factor-β2 in bovine milk during indirect heat treatments

Transforming growth factor-β2 (TGF-β2) in bovine milk was studied following heat treatments at 65 °C, 72 °C, 90 °C or 135 °C for 15 s in pilot scale indirect heating equipment. The concentrations of TGF-β2 were measured in milk, casein and whey fractions, and these values were compared with those obtained with non-heat-treated milk samples. Heat treatment had a strong effect on the growth factor and its distribution to casein and whey. Heating activated TGF-β2 from latent to immunoreactive form, as measured by immunoassay. Activation was dependent on the heating temperature and the change was total at 90 °C. TGF-β2 found in the casein fraction was in the active form, whereas the latent form of the growth factor was located in the whey fraction. TGF-β2 contents in cheese and whey were compared with that in the original milk, and these results supported the findings of the pilot scale study.