- Email: [email protected]
Plasmin and Plasminogen in Bovine Milk: A Relationship with Involution?
P l a s m i n a n d P l a s r n i n o g e n in B o v i n e Milk: A Relationship with Involution? L POLITIS, E. LACHANCE, E. BLOCK, and J. D. TURNER Department of Animal Science Macdonald College of McGill University 21,111 Lakeshore Road Ste. Anne de Bellevue Quebec H9X 1C0 Canada ABSTRACT
A total of 774 individual milk samples were collected from 66 Holstein cows between October 1987 and April 1988. Samples were analyzed for plasmin, plasminogen, and SCC. An increase in SCC from less than 250,000/ml to more than 1,000,000/ml resulted in an increase of plasmin, plasminogen, and serum albumin by 105, 74, and 140%, respectively. Plasminogen, plasmin, and serum albumin followed similar trends that are expected for components from blood that gain access to the alveolar lumen through ruptured epithelium caused by mastitis. Increased plasmin is the direct result of this process rather than an increase in activation of plasminogen to plasmin. The plasminogen to plasmin ratio supports this interpretation, being 4.7 at 250, 000 SCC/ml and 4.0 when SCC exceeded 1 million/ml. Plasmin and plasminogen concentrations were also increased during lactation to reach peak values immediately before the dry period. However, in this case, ratio of plasminogen to plasmin was 6.55 during early lactation and decreased by half to 3.29 during the latest stage, indicating that considerable activation of plasminogen to plasmin occurred during the latter part of lactation. Mammary epithelium is not compromised at this stage, as shown by low (.8 mg/ml) serum albumin concentration in milk. Two mechanisms responsible for increased milk plasmin include influx of plasmin from blood during mastiffs and
Received May 16, 1988. Accepted October 24, 1988.
1989 J Dairy Sci 72:900-906
increased activation of plasminogen as lactation progresses. INTRODUCTION
Bovine milk contains several serine proteases. Plasmin that occurs in milk together with its inactive zymogen, plasminogen, is the most significant protease in total proteolytic activity (6, 7). Plasminogen activators are found in milk (10) and convert plasminogen to plasmin. Once present, plasmin hydrolyzes ¢q-casein and [3-casein, but ~:-kasein is probably resistant to degradation (7). The extent of proteolysis affects the coagulating properties of milk (16) and also its ability to withstand processing during cheese making. As plasmin has high heat stability (4), increased concentrations can potentially present a problem to the dairy industry. Mastitic infection of the mammary gland results in increased proteolytic activity that is related to a higher concentration of plasmin (1, 2, 5, 22, 23; Politis and Ng Kwai Hang, unpublished). Stage of lactation also affects plasmin with late lactation associated with higher concentrations of plasmin (10, 19; Politis and Ng Kwai Hang, unpublished). However, these studies report correlations and not causative relationships. The incompleteness of our understanding is becoming more apparent when we try to explain why inflammation of the mammary gland or late lactation results in considerable increase of plasmin in milk. The objective of this study was to develop a model describing the plasminogen/plasmin system in bovine milk. Relationships between several factors (stage of lactation, SCC, lactation number, and season) and concentrations of plasmin and plasminogen were evaluated after exhaustive sampling. Imponam relationships were then further assessed as to their causative 900
PLASMIN AND PLASMINOGEN IN BOVINE MILK involvement in the plasmin/plasminogen system. MATERIALS AND METHODS
Individual milk samples were collected biweekly from 66 Holstein cows at the Macdonald College farm between October 1987 and April 1988. A Fossomatic cell counter (Foss Electric, Hillerod, Denmark) was used to determine SCC in milk. Milk samples were defatted and skim milk centrifuged at 100,000 x g for 1 h at 4°C to resolve the supernatant, milk serum fraction and a pellet, the casein fraction. Plasmin and plasminogen concentrations in milk serum of casein fraction were determined by the method of Korycka-Dahl et al. (10) with a slight modification (12). Briefly, this method involves assaying for plasmin activity by measuring the rate of hydrolysis of the chromogenic substrate (H-D-valyl-L-leucyl-Llysine-p-nitroanilide dihydrochloride (S-2251 Sigma Chemical Co, St Louis, MO). Formation of p-nitroanilide during cleavage of the substrate by plasmin was measured by absorbance at 405 nm. A standard curve was prepared to convert plasmin activity to plasmin concentration by plotting change in absorbance versus concentration of plasmin over the range of 0 to 3 mg/L. Concentration of serum albumin in milk was calculated from the corresponding densitometric peak area of the electrophoretic separation of whey proteins (13) using the total whey protein concentration. Whey proteins were prepared as described by Ng Kwai Hang and Kroeker (13). The ability of somatic cells to synthesize and secrete plasmin or plasminogen activators was assessed in vitro. Somatic cells were isolated from whole milk by centrifugation at 8000 x g for 10 min, followed by washing the cell pellets in Dulbecco's phosphate-buffered saline (Gibco, Grand Island, NY). Skim milk was filtered through glass wool to remove fat particles and the somatic cells were resuspended*in this skim milk to a cell density of 1,000,000/ml. Cell density was subsequently counted with a Fossomatic cell counter. Sampies were incubated at 37°C for 24 h to simulate in vitro conditions and plasmin and plasminogen concentrations were determined. The same experiment was repeated with the only
901
difference that the somatic ceils were passed through six cycles of freezing and thawing to disrupt their membranes and release intracellular components. Analysis of variance was performed to examine the effect of the environmental factors on plasmin and plasminogen concentrations in milk. Model fitted to the data included stage of lactation, SCC, season, lactation number, and cow effect as fixed classification effects and a random residual term. There were six subclasses for stage of lactation with subclasses 1, 2, 3, and 4 consisting of cows at 1st and 2nd, 3rd and 4th, 5th and 6th, and 7th and 8th mo of lactation, respectively. Subclass 5 included cows in lactation for 9 mo and subclass 6 included cows at more than 9 mo in lactation. The five subclasses for SCC were (<250,000, 250,001 to 500,000, 500,001 to 750,000, 750,001 to 1,000,000, and >1,000,000 somatic cells/ml of milk. Three subclasses for seasons: fall (October, November), winter (December, January, February), and spring (March, April) were chosen. There were five subclasses for lactation number (1, 2, 3, 4, >5). All analyses were conducted using the General Linear Model procedure of the Statistical Analysis System (26). RESULTS AND DISCUSSION
Overall mean plasmin and plasminogen concentrations and their distribution between the casein and the serum fraction in bovine milk from a total of 774 samples appear in Table 1. The average value for plasminogen in milk (.94 mg/L) is within the range reported in the literature. Richardson and Pearce (19) and Richardson (18) reported values ranging from .55 to 2.75 mg/L while Korycka-Dahl et al. (10) found an overall plasminogen concentration of 1.02 mg/L. The average plasmin conTABLE 1. Distribution of plasmin and plasminogen concentrations between the casein and the serum fraction of bovine milk. Milk fraction
Plasmin
Plasminogen Crag/L)
Serum Casein
.21 .013
SD
~
SD
.05 .003
.94 .10
.30 .02
Journal of Dairy Science Vol. 72, No. 4, 1989
902
POLITIS ET AL.
TABLE 2. Analysis of variance for factors affecting the plasmin and plasminogen system in milk. Sources of variation
df
Sum of squares Plasmin Plasminogen
Individual cow Somatic cell counts Stage of lactation Season Lactation number Residual
65 4 5 2 4 694
1.40" .40* .40* .11" .01 3.81
15.68" 32.19" 3.80* 6.15" .15 75.93
Relative importance1 Plasmin Plasminogen (%) 16.7 5.89 4.77 12.09 4.77 1.43 1.31 2.31 .00 .00 . . . . . .
1Relative importance was calculated by dividing the partial sum of the squares (type HI SS in SAS) (28) by the corrected total sum of the squares. *P<.01.
centration of .21 mg/L found in this study is higher than .1 m/L found by Korycka-Dahl et al. (10), similar to those of Richardson (18) .15 to 37 nag/L), but lower than those of RoUema et al. (20) (.7 to 2.4 mg/L). Discrepancies could be attributed to the widely different methods of measuring plasmin and plasminogen used in these studies. Also, the nature of milk being tested (origin, previous treatment) has not always been sufficiently documented to allow direct comparison of the results. Reimerdes (17) reported the presence of two serine proteases in milk, the first being plasmin and the second a thrombin-like enzyme. To evaluate the relative contribution of these enzymes pure bovine thrombin or pure plasmin (1 ~tg,/ml) was added to the reaction mixture devoid of serum. The thrombin hydrolyzed the S-2251 substrate at a rate approximately 3% compared with the 100% of plasmin. These results corroborate earlier findings by KoryckaDahl et al. (10) and Lottenberg et al. (12) who reported that thrombin hydrolyzes S-2251 at a negligible rate. Concentrations of plasmin and plasminogen were determined in the milk serum after preincubation with e-amino-n-caproic acid (EACA). The EACA is a lysine analog and is known to produce dramatic changes in the plasminogen conformation (3). This has been reported to enhance the sensitivity of plasmin assays by accelerating the rate of the conversion of plasminogen to plasmin (25). To investigate if EACA could inhibit the hydrolysis of S-2251 by plasmin, skim milk was incubated Journal of Dairy Science Vol. 72, No. 4, 1989
with the following concentrations of EACA: 2.5, 5, 10, 20, 40, 80, and 120 mM. Within the range of 2.5 to 10 mM, there was no inhibition. The hydrolysis of S-2251 by plasmin was partiaUy inhibited by EACA concentrations between 20 and 80 mM and completely inhibited at 120 mM of EACA. Plasmin determinations in this study were conducted with 2.5 mM EACA. Searls (25) reported that the use of EACA offers many advantages, the most important being that it strongly inhibits the binding of antiplasmin to plasmin by blocking lysine-binding sites in plasmin. Effects of Environmental Factors on Plasmin/Plasminogen System in Milk
Results of the analysis of variance of the variation in plasmin and plasminogen concentration due to environmental factors appear in Table 2. All factors included in the model (individual cow, stage of lactation, SCC, and season) had an effect on plasmin and plasminogen concentration in milk (P<.01). Sum of squares (SS) in Table 2 are the marginal sum of squares (type III SS in SAS) (26). In this way the effect of any factor on plasmin and plasminogen concentration in milk can be calculated with adjustments being made for the effects of all other factors in the model. Thus, the described associations are direct. Relative importance is not a measurement of the contribution of a particular effect to the total variation but one that enables the comparison of the effects relative to each other. Stage of lactation and SCC were of greater importance (4.77%)
PLASMIN AND PLASMINOGEN IN BOVINE MILK
903
The ratio of plasminogen to plasmin, which can be used as an indicator of the activation of plasminogen to plasmin, was 4.7 at Plasminogen 250,000 SCC/ml and became 4.0 when SCC 0 were more than 1,000,000/ml. Even though the actual mode of transport of plasminogen and plasmin from blood to milk is unknown, it (0 seems reasonable to assume that in mastitic quarters where the permeability of the mammary epithelium is compromised, more plasmin and plasminogen will be transferred to the 2 milk. Other plasma proteins with slightly lower t~ molecular weights than plasminogen (82,000) 3 ~- such as serum albumin and antitrypsin with a molecular weight of 70,000 and 68,000, respectively, pass through mammary epithelium (23). Philippy and McCarthy (15) estimated that at .I--I" least 80% of the albumin found in goat milk was of blood origin. Figure 1 also shows the ~:~ 0 I | relationship between SCC and serum albumin in milk. With an increase of SCC from 250,000 Somatic Cell Count, (1QS/rnll to >1,000,000 there was a concurrent rise in Figure 1. Effect of somatic cell count on A) milk serum albumin from .5 to 1.2 mg/ml. Similar plasminogen (O); and plasmin (O) concentrations and the calculated ratio of plasminogen to plasmin (11). B) Effect results were reported by Kroeker et al. (11). of somatic cell count on milk serum albumin concentra- Serum albumin and plasminogen followed simtions. ilar trends with increasing SCC. This would be expected for two components with the same origin (blood) and that gain access to the alveolar lumen through ruptured epithelium caused than season (1.31%) and lactation number in by mastiffs. Somatic cells in bovine milk are mostly explaining variation of plasmin in milk. Somatic cell count was a more important factor neutrophils and macrophages. Plasminogen ac(12.09%) than either season (2.31%) or stage of tivators from these cells (21) could potentially lactation (1.43%) in determining plasminogen activate plasminogen to plasmin and therefore decrease the ratio of plasminogen/plasmin. This variability in milk. Plasmin and plasminogen concentrations was not the case as the plasminogen to plasmin were increased with increasing SCC (Figure 1). ratio was relatively constant at 4.0 to 4.7 over a An increase of SCC from 250,000 to 750,000/ range of SCC from <250,000 to >l,000,000/ml. ml resulted in an increase of plasmin concentra- Thus, it is unlikely that plasminogen activators tion from .18 to .28 mg/L. The increase in this originating from somatic cells contribute to the range is linear, where each 250,000 increase of increase in plasmin observed in milk from masSCC/ml led to an increase in plasmin concen- titic quarters. These findings corroborate obsertration of .03 mg/L. Milk containing more than vations by Guidry (8), who reported that severe 1,000,000 SCC/ml was associated with a fur- mastitis resulted in the formation of fibrin clots ther increase in concentration of plasmin to .37 that block the ducts and prevent milk drainage rag/L. These findings are essentially in agree- of that area of the gland. Our findings are not in ment with those of others (2, 5; Politis and Ng agreement with those of Honkanen-Buzalski Kwai Hang, unpublished). The trend for plas- and Sandholm (9) or Schaar and Funke (23), minogen was similar to that of plasmin in that who reported an inhibition of the activation of the increase of SCC from 250,000/ml to more plasmin at very high SCC. This discrepancy than 1,000,000/ml led to an increase of plas- can be explained partially by the limited numminogen concentration from .85 to 1.48 mg/L. ber of milk samples in these previous studies 1.4
/
c-
Journal of Dairy Science Vol. 72, NO. 4, 1989
904
POLITIS ET AL.
and the absence of adjustments for some factors considered in this study. In some cases the increase in plasmin and plasminogen concentration in milk serum from mastitic quarters has been attributed to a redistribution of plasminogen and plasmin from the casein to the serum fraction. In our study this transfer had a minor contribution to the approximately 105 and 84% increase in plasminogen and plasmin, respectively in milks containing more than 1,000,000 when compared to that of <250,000 SCC/ml. In milk samples containing 250,000 SCC/ml, 12.5% of the plasminogen and 10.8% of the plasmin were found in the casein fraction. The corresponding figures for milk containing more than 1,000,000 SCC/ml were 9 and 7.8%, respectively. These findings are in agreement with those of Schaar and Funke (23). The presence of nonserine proteinase in milk could confound the interpretation of plasmin assay results through the hydrolysis of the chromogenic substrate. To quantitate the extent of hydrolysis, plasmin activity was inhibited with .6 IU/ml aprotinin. Addition of aprotinin reduced S-2251 hydrolysis by 95 and 92% for normal (<250,000 cell/ml) or mastitis (>1 xl06 cells/ml), respectively. The residual hydrolysis would thereby cause a slight overestimate of plasminogen and plasmin concentrations. The previously described association of SCC and plasmin and plasminogen does not establish any cause and effect relationship. On the contrary, it seems to be the outcome of increased transport of serum proteins including plasminogen and plasmin as well as somatic cells from the blood into milk due to mastitic damage. It was then thought useful to establish whether somatic cells themselves have an effect on plasminogen or plasmin in milk. Addition of somatic cells had a very minor effect on the plasmin/plasminogen system. However, a 30% increase in plasmin and a 5% decrease in plasminogen concentration was observed when milk somatic cells were lysed. This increase could be attributed to plasminogen activators that are present within some somatic cells (21). The limited extent to which intact cells activate plasminogen even after incubation at 370C for 24 h indicates that these activators remain within the cells in milk and do not effect milk plasmin concentrations. Even though somatic Journal of Dairy Science Vol. 72, No. 4, 1989
cells do not have a direct effect on plasmin or plasminogen, they may contribute to elevated proteolytic activity through higher concentrations of nonserine proteases as proposed by others (24, 27). The association of stage of lactation and the plasmin/plasminogen system in milk is shown in Figure 2. Plasmin concentration was increased from .15 mg/L in early lactation to .22 mg/L during middle lactation to reach the maximum value of .38 mg/L during the latter part of the lactation. Plasminogen was also increased during lactation, from .94 mg/L during early lactation to 1.25 mg/L during late lactation. The ratio of plasminogen to plasmin, which was 6.23 during early lactation, decreased to 3.29 during the latter part of lactation. Activation of plasminogen to plasmin, therefore, occurs as lactation progresses. Mammary epithelium integrity is not compromised at this stage, as evidenced by low (.8 mg/ml) milk serum albumin concentrations (Figure 2). Plasminogen influx from the blood is limited, and the increase in milk plasmin can be attributed to activation of milk plasminogen to plasmin. To evaluate whether this trend persists
7 "~
c-
1.0
E
6
tO
12. t-
8
n
.4~
5&
.2
4~
o
O rE
h-
I
i
i
i
i
i
B
4~
o
1:2 3"4 5"6 7"8 ; >'9 Stage of Lactation (month)
Figure 2. Effect of stage of lactation on A) milk plasminogen (O); and plasmin (0) concentrations and the calculated ratio of plasminogen to plasmin (11). B) Effect of stage of lactation on milk serum albumin concentration.
PLASMIN AND PLASMINOGEN I N BOVINE MILK
905
plasminogen concentrations between milk from cows of different lactation numbers. Our findings are not in agreement with those of Schaar and Funke (23), who reported that plasminogen tended to decrease with increasing lactation number.
1.4
1.2 ~ 1.0
CONCLUSIONS
~ ~ 4' i-
o - -
o
2 Plasmin
0
!
Fall
I
Winter
t
Spring
Figure 3. Relationship between season of the year and milk plasminogen (O) and plasmin (0) concentration.
into the dry period, milk samples were collected from cows 72 h after the cessation of milk collection. The 72-h interval was chosen because it was assumed that this would be enough time for involution process to be sufficiently advanced. Plasmin concentration was increased to .60 mg/L while plasminogen concentration increased to 1.8 mg/L, resulting in a further drop of the ratio of plasminogen to plasmin from 3.29 at the end of lactation to 3.0. We postulate that plasminogen activator secretion increases within the mammary gland as lactation progresses and results in local plasmin production. Plasmin so produced could then initiate proteolysis within the alveolus, which must occur before the tissue remodeling involved with involution can occur. Such a system is known to function in rodent mammary gland involution as described by Ossowski et al. (14). Differences in plasmin and plasminogen concentrations were found between samples from different seasons (Figure 3). Plasminogen concentrations were 44 and 20.4% higher during winter than in fall and spring, respectively. Plasmin concentration increased by 35 and 12% in the winter when compared with that of fall and spring, respectively. There was no significant (P>.05) variation in plasmin and
Our study identified factors that affect plasmin and plasminogen concentration in milk. Higher plasminogen and plasmin were associated with increasing SCC, the latter part of the lactation, and the winter months. These variations, particularly those resulting from elevated SCC, indicate that the increased plasmin in milk associated with increasing SCC is not due to an increased activation of plasminogen to plasmin but rather a result of increased appearance of plasminogen from blood to milk when the integrity of the mammary epithelium is compromised. A different pattern was observed in late lactation: the increased concentration of plasmin, with the highest value immediately before involution, is a direct result of increased activation of plasminogen to plasmin. ACKNOWLEDGMENTS
Supported in part by funds from the Natural Sciences and Engineering Research Council of Canada, Grant OGP0036728 to JDT. The authors extend thanks to Urs Kuhnlein for support and advice and Elizabeth Gill for secretarial assistance. Thanks to go the Dairy Herd Analysis Service (DHAS) for SCC determinations in their milk analysis laboratory. REFERENCES
1 Andrews, A. T. 1983. Breakdown of caseins by proteinases in bovine milks with high somatic cell counts arising from mastiffs or infusion with bacterial endotoxin. J. Dairy Res. 50:57. 2 Barry, J. G., and W. J. Donelly. 1981. Casein Compositional studies. 1I. The effect of secretory disturbance of casein composition in freshly drawn and aged bovine milk. J. Dairy Sci. 64:1038. 3 Castellino, F. J., and J. R. Powell. 1981. Human plasminogen. Methods Enzymol. 80:365. 4 Driessen, F. M., and C. B. Van der Waals. 1978. Inactivation of native milk protease by heat treatment. Neth. Milk Dairy J. 32:245. 5 Donnely, W. J., and J. G. Barry. 1983. Casein composiJournal of Dairy Science Vol. 72, No. 4, 1989
906
POLITIS ET AL.
tional studies. III. Changes in Irish milk for manufacturing and role of milk proteinases. J. Dairy Res. 50: 433. 6 Eigel, W. N. 1977. Effect of bovine plamlin on Ctsl-, [~-, and ~:-A caseins. J. Dairy Sci. 60:1399. 7 Fox, P. F. 1981. Proteinases in dairy technology. Neth. Milk Dairy J. 35:233. 8 Guidry, A. J. 1985. Mastiffs and the immune system of the mammary gland, B. L. Larson, ed. Iowa State Univ. Press, Ames. 9 Honkanen-Buzalski, T., and M. Sandhoim. 1981. Trypsin-inhibitors in mastitic milk and colostrum: correlation between trypsin-inhibitor capacity, bovine serum albumin and somatic cell counts. J. Dairy Res. 48:213. 10 Korycka-Dahl, M., B. Ribadean-Dumas, N. Chene, and J. Martal. 1983. Plasmin activity in milk. J. Dairy Sci. 66:704. 11 Kroeker, E. M., K. F. Ng-Kwai-Hang, J. F. Hayes, and J. E. Moxley. 1985. Effect of ~-lactoglobulin variant and environmental factors on variation in the detailed composition of bovine milk serum proteins. J. Dairy Sci. 68:1637. 12 Lottenherg, R,, U. Christensen, C. M. Jackson, and P. L. Coleman. 1981. Assay of coagulation proteases using peptide chromogenic and fluorogenic substrates. Methods Eazymol. 80:341. 13 Ng Kwai Hang, K. F., and E. Kroeker. 1984. Rapid separation and quantification of major caseins and whey proteins of bovine milk by polyacrylamide gel electropboresis. J. Dairy Sci. 67:3052. 14 Ossowski, L., D. Biegel, and E. Reich. 1979. Mammary plasminogen activator: correlation with involution, hormonal modulation and comparison between normal and neoplastic tissue. Cell 16:929. 15 Philippy, B. O., and R. D. McCarthy. 1979. Multi-origins
Journal of Dairy Science Vol. 72, No. 4, 1989
of milk serum albumin in the lactating goat. Biochem. Biophys. Acta 584:298. 16 Politis, I., and K. F. Ng Kwal Hang. 1988. Effects of somatic cell counts and milk composition on coagulating properties of milk. J. Dairy Sci. 71:1740. 17 Reimerdes, E. H. 1978. New results about milk serine proteinases. Neth. Milk Dairy J. 35:287. 18 Richardson, B. C. 1983. Variation of the concentration of plasmin and plasminogen in bovine milk with lactation. N.Z.J. Dairy Sci. Tecimol. 18:247. 19 Richardson, B. C., and K. N. Pearce. 1981. The determination of plasmin in dairy products. N.Z.J. Dairy Sci. Technol. 16:209. 20 RoUema, H. S., S. Visser, and I. K. Poll. 1983. Spectrophotometric assay of plantain and pla~minogen in bovine milk. Milchwissenschaft 38:214. 21 Saksella, O. 1985. Plasminogen activation and regulation of pericellular proteolysis. Biochem. Biophys. Acta 823:35. 22 Schaar, J. 1985. Plasmin activity and protease-peptone content of individual cows. J. Dairy Res. 52:369. 23 Schaar, J., and H. Funke. 1986. Effect of subclinical mastitis on milk plasminogen and plasmin compared with that on sodium, antitrypsin and N-acetyl-~i-D-glucosaminidase. J. Dai~ Res. 53:15, 24 Seaman, A. I., R. J. Verdi, D. M. Galton, and D. M. Barbano. 1988. Effect of mastiffs on proteolytic activity in bovine milk. J. Dairy Sci. 71:505. 25 Searls, O. B. 1980. An improved colodmetric assay for plasminogen activator. Anal. Biechem. 107:64. 26 Statistical Analysis System Institute Inc. 1982. SAS User's guide: statistics. Cary, NC. 27 Verdi, R. J., and D. M. Barbano. 1988. Preliminary investigation of the properties of somatic cell protcases. J. Dairy Sci. 71:534.
A total of 774 individual milk samples were collected from 66 Holstein cows between October 1987 and April 1988. Samples were analyzed for plasmin, plasminogen, and SCC. An increase in SCC from less than 250,000/ml to more than 1,000,000/ml resulted in an increase of plasmin, plasminogen, and serum albumin by 105, 74, and 140%, respectively. Plasminogen, plasmin, and serum albumin followed similar trends that are expected for components from blood that gain access to the alveolar lumen through ruptured epithelium caused by mastitis. Increased plasmin is the direct result of this process rather than an increase in activation of plasminogen to plasmin. The plasminogen to plasmin ratio supports this interpretation, being 4.7 at 250, 000 SCC/ml and 4.0 when SCC exceeded 1 million/ml. Plasmin and plasminogen concentrations were also increased during lactation to reach peak values immediately before the dry period. However, in this case, ratio of plasminogen to plasmin was 6.55 during early lactation and decreased by half to 3.29 during the latest stage, indicating that considerable activation of plasminogen to plasmin occurred during the latter part of lactation. Mammary epithelium is not compromised at this stage, as shown by low (.8 mg/ml) serum albumin concentration in milk. Two mechanisms responsible for increased milk plasmin include influx of plasmin from blood during mastiffs and
Received May 16, 1988. Accepted October 24, 1988.
1989 J Dairy Sci 72:900-906
increased activation of plasminogen as lactation progresses. INTRODUCTION
Bovine milk contains several serine proteases. Plasmin that occurs in milk together with its inactive zymogen, plasminogen, is the most significant protease in total proteolytic activity (6, 7). Plasminogen activators are found in milk (10) and convert plasminogen to plasmin. Once present, plasmin hydrolyzes ¢q-casein and [3-casein, but ~:-kasein is probably resistant to degradation (7). The extent of proteolysis affects the coagulating properties of milk (16) and also its ability to withstand processing during cheese making. As plasmin has high heat stability (4), increased concentrations can potentially present a problem to the dairy industry. Mastitic infection of the mammary gland results in increased proteolytic activity that is related to a higher concentration of plasmin (1, 2, 5, 22, 23; Politis and Ng Kwai Hang, unpublished). Stage of lactation also affects plasmin with late lactation associated with higher concentrations of plasmin (10, 19; Politis and Ng Kwai Hang, unpublished). However, these studies report correlations and not causative relationships. The incompleteness of our understanding is becoming more apparent when we try to explain why inflammation of the mammary gland or late lactation results in considerable increase of plasmin in milk. The objective of this study was to develop a model describing the plasminogen/plasmin system in bovine milk. Relationships between several factors (stage of lactation, SCC, lactation number, and season) and concentrations of plasmin and plasminogen were evaluated after exhaustive sampling. Imponam relationships were then further assessed as to their causative 900
PLASMIN AND PLASMINOGEN IN BOVINE MILK involvement in the plasmin/plasminogen system. MATERIALS AND METHODS
Individual milk samples were collected biweekly from 66 Holstein cows at the Macdonald College farm between October 1987 and April 1988. A Fossomatic cell counter (Foss Electric, Hillerod, Denmark) was used to determine SCC in milk. Milk samples were defatted and skim milk centrifuged at 100,000 x g for 1 h at 4°C to resolve the supernatant, milk serum fraction and a pellet, the casein fraction. Plasmin and plasminogen concentrations in milk serum of casein fraction were determined by the method of Korycka-Dahl et al. (10) with a slight modification (12). Briefly, this method involves assaying for plasmin activity by measuring the rate of hydrolysis of the chromogenic substrate (H-D-valyl-L-leucyl-Llysine-p-nitroanilide dihydrochloride (S-2251 Sigma Chemical Co, St Louis, MO). Formation of p-nitroanilide during cleavage of the substrate by plasmin was measured by absorbance at 405 nm. A standard curve was prepared to convert plasmin activity to plasmin concentration by plotting change in absorbance versus concentration of plasmin over the range of 0 to 3 mg/L. Concentration of serum albumin in milk was calculated from the corresponding densitometric peak area of the electrophoretic separation of whey proteins (13) using the total whey protein concentration. Whey proteins were prepared as described by Ng Kwai Hang and Kroeker (13). The ability of somatic cells to synthesize and secrete plasmin or plasminogen activators was assessed in vitro. Somatic cells were isolated from whole milk by centrifugation at 8000 x g for 10 min, followed by washing the cell pellets in Dulbecco's phosphate-buffered saline (Gibco, Grand Island, NY). Skim milk was filtered through glass wool to remove fat particles and the somatic cells were resuspended*in this skim milk to a cell density of 1,000,000/ml. Cell density was subsequently counted with a Fossomatic cell counter. Sampies were incubated at 37°C for 24 h to simulate in vitro conditions and plasmin and plasminogen concentrations were determined. The same experiment was repeated with the only
901
difference that the somatic ceils were passed through six cycles of freezing and thawing to disrupt their membranes and release intracellular components. Analysis of variance was performed to examine the effect of the environmental factors on plasmin and plasminogen concentrations in milk. Model fitted to the data included stage of lactation, SCC, season, lactation number, and cow effect as fixed classification effects and a random residual term. There were six subclasses for stage of lactation with subclasses 1, 2, 3, and 4 consisting of cows at 1st and 2nd, 3rd and 4th, 5th and 6th, and 7th and 8th mo of lactation, respectively. Subclass 5 included cows in lactation for 9 mo and subclass 6 included cows at more than 9 mo in lactation. The five subclasses for SCC were (<250,000, 250,001 to 500,000, 500,001 to 750,000, 750,001 to 1,000,000, and >1,000,000 somatic cells/ml of milk. Three subclasses for seasons: fall (October, November), winter (December, January, February), and spring (March, April) were chosen. There were five subclasses for lactation number (1, 2, 3, 4, >5). All analyses were conducted using the General Linear Model procedure of the Statistical Analysis System (26). RESULTS AND DISCUSSION
Overall mean plasmin and plasminogen concentrations and their distribution between the casein and the serum fraction in bovine milk from a total of 774 samples appear in Table 1. The average value for plasminogen in milk (.94 mg/L) is within the range reported in the literature. Richardson and Pearce (19) and Richardson (18) reported values ranging from .55 to 2.75 mg/L while Korycka-Dahl et al. (10) found an overall plasminogen concentration of 1.02 mg/L. The average plasmin conTABLE 1. Distribution of plasmin and plasminogen concentrations between the casein and the serum fraction of bovine milk. Milk fraction
Plasmin
Plasminogen Crag/L)
Serum Casein
.21 .013
SD
~
SD
.05 .003
.94 .10
.30 .02
Journal of Dairy Science Vol. 72, No. 4, 1989
902
POLITIS ET AL.
TABLE 2. Analysis of variance for factors affecting the plasmin and plasminogen system in milk. Sources of variation
df
Sum of squares Plasmin Plasminogen
Individual cow Somatic cell counts Stage of lactation Season Lactation number Residual
65 4 5 2 4 694
1.40" .40* .40* .11" .01 3.81
15.68" 32.19" 3.80* 6.15" .15 75.93
Relative importance1 Plasmin Plasminogen (%) 16.7 5.89 4.77 12.09 4.77 1.43 1.31 2.31 .00 .00 . . . . . .
1Relative importance was calculated by dividing the partial sum of the squares (type HI SS in SAS) (28) by the corrected total sum of the squares. *P<.01.
centration of .21 mg/L found in this study is higher than .1 m/L found by Korycka-Dahl et al. (10), similar to those of Richardson (18) .15 to 37 nag/L), but lower than those of RoUema et al. (20) (.7 to 2.4 mg/L). Discrepancies could be attributed to the widely different methods of measuring plasmin and plasminogen used in these studies. Also, the nature of milk being tested (origin, previous treatment) has not always been sufficiently documented to allow direct comparison of the results. Reimerdes (17) reported the presence of two serine proteases in milk, the first being plasmin and the second a thrombin-like enzyme. To evaluate the relative contribution of these enzymes pure bovine thrombin or pure plasmin (1 ~tg,/ml) was added to the reaction mixture devoid of serum. The thrombin hydrolyzed the S-2251 substrate at a rate approximately 3% compared with the 100% of plasmin. These results corroborate earlier findings by KoryckaDahl et al. (10) and Lottenberg et al. (12) who reported that thrombin hydrolyzes S-2251 at a negligible rate. Concentrations of plasmin and plasminogen were determined in the milk serum after preincubation with e-amino-n-caproic acid (EACA). The EACA is a lysine analog and is known to produce dramatic changes in the plasminogen conformation (3). This has been reported to enhance the sensitivity of plasmin assays by accelerating the rate of the conversion of plasminogen to plasmin (25). To investigate if EACA could inhibit the hydrolysis of S-2251 by plasmin, skim milk was incubated Journal of Dairy Science Vol. 72, No. 4, 1989
with the following concentrations of EACA: 2.5, 5, 10, 20, 40, 80, and 120 mM. Within the range of 2.5 to 10 mM, there was no inhibition. The hydrolysis of S-2251 by plasmin was partiaUy inhibited by EACA concentrations between 20 and 80 mM and completely inhibited at 120 mM of EACA. Plasmin determinations in this study were conducted with 2.5 mM EACA. Searls (25) reported that the use of EACA offers many advantages, the most important being that it strongly inhibits the binding of antiplasmin to plasmin by blocking lysine-binding sites in plasmin. Effects of Environmental Factors on Plasmin/Plasminogen System in Milk
Results of the analysis of variance of the variation in plasmin and plasminogen concentration due to environmental factors appear in Table 2. All factors included in the model (individual cow, stage of lactation, SCC, and season) had an effect on plasmin and plasminogen concentration in milk (P<.01). Sum of squares (SS) in Table 2 are the marginal sum of squares (type III SS in SAS) (26). In this way the effect of any factor on plasmin and plasminogen concentration in milk can be calculated with adjustments being made for the effects of all other factors in the model. Thus, the described associations are direct. Relative importance is not a measurement of the contribution of a particular effect to the total variation but one that enables the comparison of the effects relative to each other. Stage of lactation and SCC were of greater importance (4.77%)
PLASMIN AND PLASMINOGEN IN BOVINE MILK
903
The ratio of plasminogen to plasmin, which can be used as an indicator of the activation of plasminogen to plasmin, was 4.7 at Plasminogen 250,000 SCC/ml and became 4.0 when SCC 0 were more than 1,000,000/ml. Even though the actual mode of transport of plasminogen and plasmin from blood to milk is unknown, it (0 seems reasonable to assume that in mastitic quarters where the permeability of the mammary epithelium is compromised, more plasmin and plasminogen will be transferred to the 2 milk. Other plasma proteins with slightly lower t~ molecular weights than plasminogen (82,000) 3 ~- such as serum albumin and antitrypsin with a molecular weight of 70,000 and 68,000, respectively, pass through mammary epithelium (23). Philippy and McCarthy (15) estimated that at .I--I" least 80% of the albumin found in goat milk was of blood origin. Figure 1 also shows the ~:~ 0 I | relationship between SCC and serum albumin in milk. With an increase of SCC from 250,000 Somatic Cell Count, (1QS/rnll to >1,000,000 there was a concurrent rise in Figure 1. Effect of somatic cell count on A) milk serum albumin from .5 to 1.2 mg/ml. Similar plasminogen (O); and plasmin (O) concentrations and the calculated ratio of plasminogen to plasmin (11). B) Effect results were reported by Kroeker et al. (11). of somatic cell count on milk serum albumin concentra- Serum albumin and plasminogen followed simtions. ilar trends with increasing SCC. This would be expected for two components with the same origin (blood) and that gain access to the alveolar lumen through ruptured epithelium caused than season (1.31%) and lactation number in by mastiffs. Somatic cells in bovine milk are mostly explaining variation of plasmin in milk. Somatic cell count was a more important factor neutrophils and macrophages. Plasminogen ac(12.09%) than either season (2.31%) or stage of tivators from these cells (21) could potentially lactation (1.43%) in determining plasminogen activate plasminogen to plasmin and therefore decrease the ratio of plasminogen/plasmin. This variability in milk. Plasmin and plasminogen concentrations was not the case as the plasminogen to plasmin were increased with increasing SCC (Figure 1). ratio was relatively constant at 4.0 to 4.7 over a An increase of SCC from 250,000 to 750,000/ range of SCC from <250,000 to >l,000,000/ml. ml resulted in an increase of plasmin concentra- Thus, it is unlikely that plasminogen activators tion from .18 to .28 mg/L. The increase in this originating from somatic cells contribute to the range is linear, where each 250,000 increase of increase in plasmin observed in milk from masSCC/ml led to an increase in plasmin concen- titic quarters. These findings corroborate obsertration of .03 mg/L. Milk containing more than vations by Guidry (8), who reported that severe 1,000,000 SCC/ml was associated with a fur- mastitis resulted in the formation of fibrin clots ther increase in concentration of plasmin to .37 that block the ducts and prevent milk drainage rag/L. These findings are essentially in agree- of that area of the gland. Our findings are not in ment with those of others (2, 5; Politis and Ng agreement with those of Honkanen-Buzalski Kwai Hang, unpublished). The trend for plas- and Sandholm (9) or Schaar and Funke (23), minogen was similar to that of plasmin in that who reported an inhibition of the activation of the increase of SCC from 250,000/ml to more plasmin at very high SCC. This discrepancy than 1,000,000/ml led to an increase of plas- can be explained partially by the limited numminogen concentration from .85 to 1.48 mg/L. ber of milk samples in these previous studies 1.4
/
c-
Journal of Dairy Science Vol. 72, NO. 4, 1989
904
POLITIS ET AL.
and the absence of adjustments for some factors considered in this study. In some cases the increase in plasmin and plasminogen concentration in milk serum from mastitic quarters has been attributed to a redistribution of plasminogen and plasmin from the casein to the serum fraction. In our study this transfer had a minor contribution to the approximately 105 and 84% increase in plasminogen and plasmin, respectively in milks containing more than 1,000,000 when compared to that of <250,000 SCC/ml. In milk samples containing 250,000 SCC/ml, 12.5% of the plasminogen and 10.8% of the plasmin were found in the casein fraction. The corresponding figures for milk containing more than 1,000,000 SCC/ml were 9 and 7.8%, respectively. These findings are in agreement with those of Schaar and Funke (23). The presence of nonserine proteinase in milk could confound the interpretation of plasmin assay results through the hydrolysis of the chromogenic substrate. To quantitate the extent of hydrolysis, plasmin activity was inhibited with .6 IU/ml aprotinin. Addition of aprotinin reduced S-2251 hydrolysis by 95 and 92% for normal (<250,000 cell/ml) or mastitis (>1 xl06 cells/ml), respectively. The residual hydrolysis would thereby cause a slight overestimate of plasminogen and plasmin concentrations. The previously described association of SCC and plasmin and plasminogen does not establish any cause and effect relationship. On the contrary, it seems to be the outcome of increased transport of serum proteins including plasminogen and plasmin as well as somatic cells from the blood into milk due to mastitic damage. It was then thought useful to establish whether somatic cells themselves have an effect on plasminogen or plasmin in milk. Addition of somatic cells had a very minor effect on the plasmin/plasminogen system. However, a 30% increase in plasmin and a 5% decrease in plasminogen concentration was observed when milk somatic cells were lysed. This increase could be attributed to plasminogen activators that are present within some somatic cells (21). The limited extent to which intact cells activate plasminogen even after incubation at 370C for 24 h indicates that these activators remain within the cells in milk and do not effect milk plasmin concentrations. Even though somatic Journal of Dairy Science Vol. 72, No. 4, 1989
cells do not have a direct effect on plasmin or plasminogen, they may contribute to elevated proteolytic activity through higher concentrations of nonserine proteases as proposed by others (24, 27). The association of stage of lactation and the plasmin/plasminogen system in milk is shown in Figure 2. Plasmin concentration was increased from .15 mg/L in early lactation to .22 mg/L during middle lactation to reach the maximum value of .38 mg/L during the latter part of the lactation. Plasminogen was also increased during lactation, from .94 mg/L during early lactation to 1.25 mg/L during late lactation. The ratio of plasminogen to plasmin, which was 6.23 during early lactation, decreased to 3.29 during the latter part of lactation. Activation of plasminogen to plasmin, therefore, occurs as lactation progresses. Mammary epithelium integrity is not compromised at this stage, as evidenced by low (.8 mg/ml) milk serum albumin concentrations (Figure 2). Plasminogen influx from the blood is limited, and the increase in milk plasmin can be attributed to activation of milk plasminogen to plasmin. To evaluate whether this trend persists
7 "~
c-
1.0
E
6
tO
12. t-
8
n
.4~
5&
.2
4~
o
O rE
h-
I
i
i
i
i
i
B
4~
o
1:2 3"4 5"6 7"8 ; >'9 Stage of Lactation (month)
Figure 2. Effect of stage of lactation on A) milk plasminogen (O); and plasmin (0) concentrations and the calculated ratio of plasminogen to plasmin (11). B) Effect of stage of lactation on milk serum albumin concentration.
PLASMIN AND PLASMINOGEN I N BOVINE MILK
905
plasminogen concentrations between milk from cows of different lactation numbers. Our findings are not in agreement with those of Schaar and Funke (23), who reported that plasminogen tended to decrease with increasing lactation number.
1.4
1.2 ~ 1.0
CONCLUSIONS
~ ~ 4' i-
o - -
o
2 Plasmin
0
!
Fall
I
Winter
t
Spring
Figure 3. Relationship between season of the year and milk plasminogen (O) and plasmin (0) concentration.
into the dry period, milk samples were collected from cows 72 h after the cessation of milk collection. The 72-h interval was chosen because it was assumed that this would be enough time for involution process to be sufficiently advanced. Plasmin concentration was increased to .60 mg/L while plasminogen concentration increased to 1.8 mg/L, resulting in a further drop of the ratio of plasminogen to plasmin from 3.29 at the end of lactation to 3.0. We postulate that plasminogen activator secretion increases within the mammary gland as lactation progresses and results in local plasmin production. Plasmin so produced could then initiate proteolysis within the alveolus, which must occur before the tissue remodeling involved with involution can occur. Such a system is known to function in rodent mammary gland involution as described by Ossowski et al. (14). Differences in plasmin and plasminogen concentrations were found between samples from different seasons (Figure 3). Plasminogen concentrations were 44 and 20.4% higher during winter than in fall and spring, respectively. Plasmin concentration increased by 35 and 12% in the winter when compared with that of fall and spring, respectively. There was no significant (P>.05) variation in plasmin and
Our study identified factors that affect plasmin and plasminogen concentration in milk. Higher plasminogen and plasmin were associated with increasing SCC, the latter part of the lactation, and the winter months. These variations, particularly those resulting from elevated SCC, indicate that the increased plasmin in milk associated with increasing SCC is not due to an increased activation of plasminogen to plasmin but rather a result of increased appearance of plasminogen from blood to milk when the integrity of the mammary epithelium is compromised. A different pattern was observed in late lactation: the increased concentration of plasmin, with the highest value immediately before involution, is a direct result of increased activation of plasminogen to plasmin. ACKNOWLEDGMENTS
Supported in part by funds from the Natural Sciences and Engineering Research Council of Canada, Grant OGP0036728 to JDT. The authors extend thanks to Urs Kuhnlein for support and advice and Elizabeth Gill for secretarial assistance. Thanks to go the Dairy Herd Analysis Service (DHAS) for SCC determinations in their milk analysis laboratory. REFERENCES
1 Andrews, A. T. 1983. Breakdown of caseins by proteinases in bovine milks with high somatic cell counts arising from mastiffs or infusion with bacterial endotoxin. J. Dairy Res. 50:57. 2 Barry, J. G., and W. J. Donelly. 1981. Casein Compositional studies. 1I. The effect of secretory disturbance of casein composition in freshly drawn and aged bovine milk. J. Dairy Sci. 64:1038. 3 Castellino, F. J., and J. R. Powell. 1981. Human plasminogen. Methods Enzymol. 80:365. 4 Driessen, F. M., and C. B. Van der Waals. 1978. Inactivation of native milk protease by heat treatment. Neth. Milk Dairy J. 32:245. 5 Donnely, W. J., and J. G. Barry. 1983. Casein composiJournal of Dairy Science Vol. 72, No. 4, 1989
906
POLITIS ET AL.
tional studies. III. Changes in Irish milk for manufacturing and role of milk proteinases. J. Dairy Res. 50: 433. 6 Eigel, W. N. 1977. Effect of bovine plamlin on Ctsl-, [~-, and ~:-A caseins. J. Dairy Sci. 60:1399. 7 Fox, P. F. 1981. Proteinases in dairy technology. Neth. Milk Dairy J. 35:233. 8 Guidry, A. J. 1985. Mastiffs and the immune system of the mammary gland, B. L. Larson, ed. Iowa State Univ. Press, Ames. 9 Honkanen-Buzalski, T., and M. Sandhoim. 1981. Trypsin-inhibitors in mastitic milk and colostrum: correlation between trypsin-inhibitor capacity, bovine serum albumin and somatic cell counts. J. Dairy Res. 48:213. 10 Korycka-Dahl, M., B. Ribadean-Dumas, N. Chene, and J. Martal. 1983. Plasmin activity in milk. J. Dairy Sci. 66:704. 11 Kroeker, E. M., K. F. Ng-Kwai-Hang, J. F. Hayes, and J. E. Moxley. 1985. Effect of ~-lactoglobulin variant and environmental factors on variation in the detailed composition of bovine milk serum proteins. J. Dairy Sci. 68:1637. 12 Lottenherg, R,, U. Christensen, C. M. Jackson, and P. L. Coleman. 1981. Assay of coagulation proteases using peptide chromogenic and fluorogenic substrates. Methods Eazymol. 80:341. 13 Ng Kwai Hang, K. F., and E. Kroeker. 1984. Rapid separation and quantification of major caseins and whey proteins of bovine milk by polyacrylamide gel electropboresis. J. Dairy Sci. 67:3052. 14 Ossowski, L., D. Biegel, and E. Reich. 1979. Mammary plasminogen activator: correlation with involution, hormonal modulation and comparison between normal and neoplastic tissue. Cell 16:929. 15 Philippy, B. O., and R. D. McCarthy. 1979. Multi-origins
Journal of Dairy Science Vol. 72, No. 4, 1989
of milk serum albumin in the lactating goat. Biochem. Biophys. Acta 584:298. 16 Politis, I., and K. F. Ng Kwal Hang. 1988. Effects of somatic cell counts and milk composition on coagulating properties of milk. J. Dairy Sci. 71:1740. 17 Reimerdes, E. H. 1978. New results about milk serine proteinases. Neth. Milk Dairy J. 35:287. 18 Richardson, B. C. 1983. Variation of the concentration of plasmin and plasminogen in bovine milk with lactation. N.Z.J. Dairy Sci. Tecimol. 18:247. 19 Richardson, B. C., and K. N. Pearce. 1981. The determination of plasmin in dairy products. N.Z.J. Dairy Sci. Technol. 16:209. 20 RoUema, H. S., S. Visser, and I. K. Poll. 1983. Spectrophotometric assay of plantain and pla~minogen in bovine milk. Milchwissenschaft 38:214. 21 Saksella, O. 1985. Plasminogen activation and regulation of pericellular proteolysis. Biochem. Biophys. Acta 823:35. 22 Schaar, J. 1985. Plasmin activity and protease-peptone content of individual cows. J. Dairy Res. 52:369. 23 Schaar, J., and H. Funke. 1986. Effect of subclinical mastitis on milk plasminogen and plasmin compared with that on sodium, antitrypsin and N-acetyl-~i-D-glucosaminidase. J. Dai~ Res. 53:15, 24 Seaman, A. I., R. J. Verdi, D. M. Galton, and D. M. Barbano. 1988. Effect of mastiffs on proteolytic activity in bovine milk. J. Dairy Sci. 71:505. 25 Searls, O. B. 1980. An improved colodmetric assay for plasminogen activator. Anal. Biechem. 107:64. 26 Statistical Analysis System Institute Inc. 1982. SAS User's guide: statistics. Cary, NC. 27 Verdi, R. J., and D. M. Barbano. 1988. Preliminary investigation of the properties of somatic cell protcases. J. Dairy Sci. 71:534.