Production of Various Forms of Plasminogen Activator and Plasminogen Activator Inhibitor by Cultured Mammary Epithelial Cells C. E. HEEGARD,' J. H. WHITE, B. ZAVIZION, J. D. TURNER? AND 1. POLITIS3 Department of Animal and Food Sciences University of Vermont Burlington 05405 ABSTRACT
We examined amounts and types of plasminogen activator and plasminogen activator inhibitor produced by cultured bovine mammary epithelial cells. The MAC-T and two other mammary epithelial cell lines, MACT-W1 and MACT-UV2 derived from the parental MAC-T cells by subcloning, were used as model systems. Cells were cultured in a medium free of serum and protein. Data showed that MACT-W2 cells produced 6.2 and 17.2% more plasminogen activator than MACT-W1 and parental MAC-T cells, respectively. Addition of amiloride, a specific urokinaseplasminogen activator inhibitor, dramatically decreased the activity in the culture medium of parental and subclonal lines, indicating that urokinase-plasminogen activator was present. Zymography revealed the presence of urokinaseplasminogen activator with an approximate molecular mass of 50,000 kDa in the culture medium of parental MAC-T cells. The culture medium of the subclonal lines contained urokinaseplasminogen activator and tissueplasminogen activator with approximate molecular masses of 50,000 and 75,000 kDa, respectively. Complexes of both types of plasminogen activators with plasminogen activator-inhibitor-1 were
detected in the culture medium of subclonal lines. (Key words: plasminogen activator, plasminogen activator-inhibitor-1, epithelial cells, involution)
Abbreviation key: FBS = fetal bovine serum, PA = plasminogen activator, PAI-1 = PA inhibitor-1, SFPF = free of serum and protein, t-PA = tissue PA, u-PA = urokinase PA. INTRODUCTION
The evidence is growing that plasminogen activator (PA) expression plays a prominent role during involution of the mammary gland postlactation (3, 12, 13, 16). In rodents, mammary gland involution involves destruction of the basement membrane and complete degeneration of secretory epithelium (12, 16). Although direct evidence is still missing, the following observations suggest a regulatory role for PA produced by mouse mammary epithelial cells in these processes: 1) PA production dramatically increases shortly after involution is initiated, and 2) enzyme synthesis is inversely correlated with the expression of differentiated phenotype that is characteristic of lactation; however, PA production is upregulated by hormones that promote involution (12). Major differences exist in studies (1) of lactation between highly inbred laboratory animals and dairy cows. Postlactational involution in dairy cows is different from that in rodents in that the events are less dramatic (7, 11). The PA activity in the mammary gland Received February 4, 1994. increased 2 to 4 d following initiation of postAccepted May 9, 1994. 'Department of Molecular Biology, University of lactational involution or in mammary secreAarhus, Aarhus C. Denmark. tions during the dry period (6, 13). However, *Department of Animal Science, McGill University, the precise role of PA in these events has not Ste. Anne de Bellevue. PQ, Canada, H9X 3V9. been established. Furthermore, the cellular ori3Reprints.
HEEGARD ET AL.
gin of PA in bovine milk and mammary extracts is not known. Bovine milk contains many types of PA. Several studies reported the presence of urokinase PA (u-PA) (4, 5) and tissue PA (t-PA) associated with milk somatic cells (19). In addition, Lu and Nielsen (9) reported the presence of five proteins in bovine milk that are capable of activating plasminogen with molecular masses of 93, 57, 42, 35, and 27 kDa. Recently, the complete cDNA sequences of bovine U-PA and t-PA were determined. The derived AA sequences coding for the mature proteins were 413 and 566 AA for u-PA and t-PA, respectively. Homologies of bovine PA with human PA were 75% (P. Ravn, Institute of Molecular Biology, Aarhus University, 1994, personal communication). Heegard et al. (4, 5 ) reported that the major type of PA associated with the casein micelles was t-PA and that they detected complexes of t-PA with PA inhibitor-1 (PAI-1). The presence of many forms of PA in bovine milk has led investigators to postulate that milk PA may originate, at least in part, in mammary epithelial cells. Earlier reports (14, 17) suggested that PA was produced by bovine mammary epithelial cells. However, those workers (14, 17) cultured the cells in the presence of fetal bovine serum (FBS),which is an abundant source of PA. To the best of our knowledge, no direct evidence exists of PA production by bovine mammary epithelial cells cultured in a serum-free medium. The objective of this study was to examine the quantity and type of PA and PA1 produced by cultured bovine mammary epithelial cells. This study provides the basis for future studies to understand the hormonal regulation of PA production and, ultimately, the role of this enzyme in involution of the bovine mammary gland. MATERIALS AND METHODS
The origin of the MAC-T mammary epithelial cell line was described previously (8). The MAC-T cells were produced from primary bovine mammary epithelial cells by stable transfection with a plasmid bearing the sequence for SV-40large T-antigen (8). Frozen Journal of Dairy Science Vol. 77, No. 10, 1994
aliquots of cells in their 20th passage were transported to our laboratory from McGill University (Montreal, PQ, Canada) and subsequently utilized in the experiments described herein. For subcloning, parental MAC-T cells were seeded on 60-mm plastic tissue culture dishes (Coming Glass Works, Corning, NY)at a density of approximately 30 cells per dish. Cells were cultivated in medium consisting of RPMI-1640 (Sigma Chemical Co., St. Louis, MO), complete Dulbecco’s Modified Eagle’s Medium (Sigma Chemical Co.), and Iscove’s Modified Dulbecco’s Medium (Sigma Chemical Co.), 1:1:1, (voYvoVvol), supplemented with 10% FBS (Sigma Chemical Co.), penicillin (100 ILJ/ml), and streptomycin (100 pg/ml). Cultures were maintained at 37°C in a humidified (loo%), 5 % C02 incubator. Cells were cultivated up to 6 to 9 d. Colonies with different growth characteristics were selected for subcloning. Cells were detached following addition of versene (.02% EDTA; Sigma Chemical Co.) in Hanks balanced salt solution (Sigma Chemical Co.) without Ca and Mg, followed by addition of versene plus .Ol% trypsin (Sigma Chemical Co.) in Hanks balanced salt solution for 2 to 3 min. This process allowed detachment of individual colonies as units. Cells from the selected colony were manually harvested using a micromanipulator (Leitz, Bremme, Germany) fitted with a glass pipette and transferred to another tissue culture dish. Trypsinization of transferred cells was allowed to proceed for an additional 12 min to yield a suspension of single cells. Cells were visualized with a Nikon (Nippon Kogaku K.K., Tokyo, Japan) inverted microscope. Homogeneity of cultures was verified using morphological characteristics and additional cloning from colonies. Eight different clonal lines were derived from the parental MAC-T cells. Two of the eight clonal lines were utilized for the PA studies; the two cell lines referred to as MACT-W 1 and MACT-UV2 represented extreme differences in growth characteristics. Cell Culture Procedures
Mammary epithelial cells @rental MAC-T and subclonal lines MACT-Wl and MACTW 2 ) were cultivated in hybridoma culture medium (Sigma Chemical Co.) that was free of
PLASMINOGEN ACTIVATORS AND EPITHELIAL CELLS
serum and protein (SFPF). Cells were maintained at 37°C in a 5% C02 incubator for 15 passages prior to use to allow an adaptation for these cells in SFPF culture medium. Approximately 5 x lo4 cells/cm2 were seeded on 60-mm plastic tissue culture dishes in SFPF medium containing .l% FBS for approximately 12 h. The small amount of FBS (1% of that normally used) was added, for this limited time, to neutralize trypsin and to allow seeding of the cells. Following a 12-h incubation, cells were allowed to grow in SFPF culture medium without FBS for up to 12 d. Medium was changed every 3 d. The medium from the last 3 d in culture (d 10, 11, and 12) was collected and stored at -20°C until further PA analysis.
system utilized the PA that is present in the culture medium to convert exogenously supplied plasminogen to active plasmin. Plasmin, so produced, is subsequently allowed to attack the chromogenic substrate Val-Leu-Lys-pnitroaniline adjacent to Lys and liberate the free chromophore p-nitroaniline. In this system, changes in color are directly related to plasmin concentrations and, therefore, indirectly to PA activity. Assays were performed in 250 pl of 100 mM Tris buffer (pH 8.0) containing plasminogen (50 pg/ml of plasminogen; American Diagnostics, Greenwich, CT), .6 mM Val-LeuLys-p-nitroanilide (V-7221; Sigma Chemical Co.), and 1 to 5 p1 of culture medium. Preliminary experiments indicated that PA activity was maximum at .6 mM of V-7221, 50 pg/ml of exogenous plasminogen and pH 8.0, and Cell Characterization these conditions were maintained throughout Growth characteristics of parental and sub- all subsequent assays. The reaction mixture clonal lines were compared. Doubling times was incubated for 3 h, and absorbance at 405 were estimated after 1 to 1.5 x lo4 cells per nm was measured at 30-min intervals using a well were seeded in 24-well multiwells (Corn- microtiter plate (Bio-Tec Instruments, Burlinging Glass Works). Cells were grown on plastic ton, VT). The rate of p-nitroaniline formation and incubated for 20 to 30 d in SFPF culture was calculated from the linear part of the curve medium, which was changed every 3 d. Cells for absorbance versus time. A sample without were maintained at 37' in a 5 % C02 incubator. plasminogen served as a control. Preliminary Cells were removed by complete trypsinization experiments indicated that PA activity was lin(12 min at 37'C) every 24 h. Total cells were ear for up to 3 h of incubation and between 0 counted using a hemocytometer. and 5 pl of sample volume. Colony-forming abilities of subclonal and parental MAC-T cells in SFPF medium were Effect of Fibrin and Amiloride measured after approximately 500 cells were on PA Activity seeded per 60-mm tissue culture dish. Cells The PA activity in culture medium was were allowed to grow on plastic for 2 wk. determined in the presence or absence of fibrin Cells were fixed with cold (-2O'C) methanol (20 pglml) or amiloride (1 mM). Preliminary for 5 min, followed by staining with .25% crystal violet (Sigma Chemical Co.) in Hanks experiments showed that fibrin and amiloride, balanced salt solution. Colonies of cells were added in these concentrations, optimally afvisualized with a Nikon inverted microscope, fected PA activity. All other details are as described. and cell numbers were recorded. The presence of cytokeratin in parental and subclonal lines was examined according to our Zymography previously published procedures (8, 21). Zymography was employed to detect the presence and the type of PA in the medium of Determination of PA Activity parental and subclonal lines. Details of zymogin Culture Medium raphy have been described earlier (2, 4, 5). A colorimetric assay, previously used to Briefly, proteins present in culture medium determine PA activity in milk somatic cells were separated by SDS-PAGE under non(19), was validated to measure PA activity reducing conditions that were carried out in present in the SFPF culture medium. The assay 10% polyacrylamide slab gels with prestained Journal of Dairy Science Vol. 77, No. 10, 1994
HEEGARD ET AL
molecular mass markers (Sigma Chemical Co.) in adjacent lanes. These markers were used to estimate the approximate molecular mass of the various PA forms present in the culture medium of mammary epithelial cells. The PA was localized in the gels by layering the SDSPAGE gels over agarose gels containing fibrin and plasminogen at 37'C (4, 5). The principle of the assay system is that PA diffuses into the agarose gels and activates plasminogen, resulting in visible lysis zones because of the action of plasmin. To identify the type of PA (u-PA vs. t-PA) present in the SFPF culture medium of parental and subclonal lines, zymography was repeated but with the addition in the agarose gel of .1 mg/ml of anti-human t-PA rabbit IgG or 2 mM of the u-PA specific inhibitor, amiloride. All details regarding this antibody generation were described by Nielsen et al. (10). This antibody can recognize bovine t-PA, but not uPA (4, 5). For control purposes, gels without added plasminogen were used to evaluate the presence of plasminogen-independent proteolytic activity. All other details of this system have been described earlier (2, 4, 5).
TABLE 1 . Comparison of important properties, mean doubling time, mean colony-forming ability, and presence of cytokeratin between three mammary epithelial cell lines: MAC-T, MACT-UVl, and MACT-UV2.I
Production of PA
Differences in PA activity in the culture medium of MAC-T, MACT-W 1, and MACTW 2 and differences between PA activity in the presence or absence of fibrin or amiloride were evaluated using Student's t test (P < .OS).
The PA activity was detected in culture medium from all cell lines (Table 2). However, subclonal lines tended to produce more PA than the parental MAC-T cells. The MACTW 2 cells produced 6.2 and 17.2% more PA
MAC-T MACT-UVl MACT-UV2
45 47 76
40 25 36.5
Cytokeratin Positive Positive Positive
IStandard errors were <3% of the mean.
RESULTS AND DISCUSSION Mammary Epithelial Cell Lines
Table 1 presents a summary of important characteristics of the parental MAC-T cells and the two subclonal lines, MACT-W1 and MACT-W2. Eight subclonal lines were generated in our laboratory, but only two were utilized in this experiment. These lines were selected based on differences in growth characteristics: MACT-W 1 grew faster than MACTW 2 . Table 1 shows that parental and subclonal lines are positive for cytokeratin peptide Immunoprecipitation of PAI Complexes 14, confirming the epithelial nature of parental To immunoprecipitate PA/PAI- 1 com- and subclonal lines. The extensive network of plexes, 200 pl of culture medium in which cytokeratins is the hallmark of mammary MAC-T, MACT-W 1, or MACT-W 2 were epithelial cells (8). Parental and subclonal lines cultivated were mixed with 0 or 5 pg of puri- were different in their ability to grow in a fied anti-human PAI-1 rabbit IgG. This anti- SFPF culture medium. Doubling times of body recognizes bovine PAI-1 (4, 5). After parental MAC-T, MACT-W1, and MACTincubation for 16 h at 4'C, 25 p1 of porcine W 2 were 45, 47, and 76.3 h, respectively. anti-rabbit IgG (Dako, Glostrup, Denmark) Colony-forming abilities of parental MAC-T, were added to facilitate precipitation. The mix- MACT-Wl, and MACT-W2 were 40, 25, ture was incubated for an additional 16 h at and 36.5%, respectively. Subcloning of the 4°C. Immunocomplexes were sedimented by MAC-T cells has been successful (21); three centrifugation at 5000 x g for 30 min. The subclonal lines of MAC-T cells developed supernatant (50 pl) of the previous centrifuga- with differences in growth properties, chrotion step was subjected to 10% SDS-PAGE, mosomal profile, and IGF-I receptor dynamics followed by zymography. (21).
Journal of Dairy Science Voi. 77, No. 10. 1994
PLASMINOGEN ACTIVATORS AND EPITHELIAL CELLS TABLE 2. Comparison of plasminogen activator (PA) activity, determined as the rate of p-nitroanilide formation, in the culture medium of three mammary epithelial cell lines: MAC-T, MACT-UVl, and MACT-UV2. Cell line
PA Activity1 (change in absorbancd~) X SE .05P ,003 0.64b .003 .M8C .003
TABLE 3. Effect of fibrin (20 pglml) and amiloride (1 mM) on plasminogen activator (PA), determined as the rate of p-nitroanilide formation in the culture medium of three mammary epithelial cell lines: MAC-T, MACT-UV1, and MACT-WZ. PA Activity'
MAC-T MACT-W 1 MAC-UV2
alb.CMeans within a column with different superscript letters differ (P < .05). lResults are means of 15 independent determinations.
than MACT-UVl and parental MAC-T cells, respectively. These data are the first to indicate that bovine mammary epithelial cells, when cultured in a SFPF medium, produce PA. Previous studies (14, 17) reported PA production by mammary epithelial cells, cultured in a serum-supplemented medium. The drawback of those studies was that MAC-T cells were cultured in the presence of FBS, which is an abundant source of PA. Characterization of PA
The effect of fibrin and amiloride on PA activity in the culture medium was examined. Two types of PA have been classified into two distinct groups: t-PA and u-PA. These PA types have different structural properties, immunochemical specificities, and sensitivity to fibrin and amiloride (3, 4, 5, 15, 18, 20). Unlike u-PA, t-PA is preferentially activated in the presence of fibrin fragments (16, 20) through binding of both activator and plasminogen to fibrin to form a cyclic complex. However, amiloride inhibits the activity of uPA, but not t-PA (5, 18). Table 3 shows that PA activity did not increase in the culture medium of MAC-T cells in the presence of fibrin. Control experiments showed that addition of fibrin increased t-PA activity sevenfold but had no effect on uPA activity (data not shown). These results, collectively, indicate that t-PA was not present in the culture medium of MAC-T cells. Activity of PA detected in the culture medium of MAC-T cells was reduced sixfold (P< .01) in
-X (change-X in absorbanceh) -X Control Fibrin Amiloride
.W ,059s .OIOb
a-brMeans within a column with different superscripts differ (P e .05). *Results am means of 5 independent determinations. Standard errors were 6% of the mean.
the presence of 1 mM amiloride, which indicates that u-PA was present in the culture medium of MAC-T cells. Effects of fibrin and amiloride on PA activity in the medium of subclonal lines (MACT-UV1 and MACT-UV2) were examined. Table 3 shows that addition of fibrin caused a small but significant (P < .OS) increase in PA activity in culture media from both cell types, indicating the presence of low concentrations of t-PA. Amiloride inhibited PA activity in media from both cell types (Table 3), indicating the presence of u-PA. These results, collectively, indicate that parental MAC-T cells produced only u-PA, but subclonal lines produced u-PA and t-PA. Further experiments verified these findings. Zymography
Control experiments showed no PA activity in plasminogen-free gels (Figure l), indicating that all of the proteolytic activities in Figure 2 were dependent on plasminogen. Zymography in plasminogen-supplemented gels revealed the presence of various bands of PA activity. The PA activity in the culture medium of parental and subclonal lines was mainly associated with a very intense lysis area that corresponded to a molecular mass of 50,000 kDa (Figure 2). However, additional lysis areas were less intense in the culture medium from the subclonal lines, but not in the medium of parental MACJournal of Daily Science Vol. 77, No. 10, 1994
HEEGARD ET AL.
affected when anti-human t-PA IgG was added to the gel. Heegard et al. (4, 5 ) demonstrated that this antibody inhibited the activity of bovine t-PA. Third, the main form of u-PA in bovine milk has a molecular mass of 50,000 kDa (4, 5 , 17). 125 kDa The areas of activity with molecular mass of 75,000 and 130,000 (Figure 2, A and B) are 88 kDa attributed to t-PA. The activity associated with the 75,000- and 130,000-kDa bands com65 kDa pletely disappeared when anti-human t-PA IgG was added to the gel (Figure 2C). Furthermore, 5 6 kDa the intensity of both bands (75,000 and 130,000 kDa) remained unaffected when amiloride was added to the gel (Figure 2B). The 130,000-kDa PA band can actually be observed better in the presence of amiloride 38kDa (Figure 2B). Heegard et al. (4) demonstrated that t-PA was present in bovine milk and had a molecular mass of 75,000 kDa. The 85,000-kDa PA band (Figure 2, A and C) apparently represents an enzymatically active form of u-PA. This suggestion is supported by the complete disappearance of its 1 2 3 activity when amiloride, a specific u-PA inhibitor, was added to the gel (Figure 2B). Furthermore, the intensity of the band remained unFigure. 1. Zymographic analysis of plasminogen activator (PA) activity in the culture medium of three mammary affected when anti-human t-PA IgG was added epithelial cell lines: MAC-T, MACT-UVI, MACT-W2. to the gel (Figure 2C). Conversion of the enzyCells were cultured for up to 12 d. Medium was collected matically active two-chain u-PA with an apduring d 10. 11, and 12 in culture and was subjected to proximate molecular mass of 50,000 kDa to an SDS-PAGE followed by zymography performed at standard conditions with the only difference being the omis- enzymatically active low molecular mass form sion of plasminogen from the agarose-fibrin gels. Lanes 1, of u-PA with an approximate molecular mass 2, and 3 were loaded with 50 p1 of medium in which of 30,000kDa has been described in several MAC-T, MACT-UVl, and MACT-UV2 cells were cul- systems (3, 16). The u-PA form with a low tured, respectively. The positions of the molecular mass molecular mass of 30,000 kDa has been demarkers are indicated to the left. tected in bovine milk (9, 15, 19) and in culture medium of bovine kidney cells (5). Because the PAI-1 has an approximate molecular mass of T cells. These lysis areas corresponded to ap- 50,000 kDa (16), the molecular mass of 85,000 proximate molecular mass of 75,000, 85,000, kDa might represent a complex of the low molecular mass form of u-PA with PAI-1. 120,000, and 130,000 kDa (Figure 2). The areas of activity with molecular mass Conclusions await further experimentation. Strong evidence indicates that the of 50,000 and 120,000 kDa (Figure 2, A and C) are attributed to u-PA. This conclusion is 120,000-kDa form in the medium of subclonal supported by three pieces of evidence. First, lines (Figure 2, A and C)represents a complex the activity associated with the 50,000-kDa of u-PAlPAI-1. Figure 3 shows lack of lysis in band was quenched, but the activity associated zymography (lanes 2 and 3) when the culture with the 120,000-kDa band completely disap- medium containing the complex was incubated peared, when amiloride was added to the gel with anti-human PAI-1 IgG. The ability of this (Figure 2B). Amiloride is a specific u-PA in- antibody to recognize bovine PAI-1 has been hibitor (4, 16). Second, the intensity of both established (4,5). The inhibitor-activator combands (50,000 and 120,000 kDa) remained un- plex is partly dissociated upon contact with Journal of Dairy Science Vol. 77. No. 10, 1994
PLASMINOGEN ACTIVATORS AND EPITHELIAL CELLS
-u-pA/I *I -1
6 5 m*
Figure 2. Zymographic analysis of plasminogen activator (PA) activity in the culture medium of h e mammary epithelial cell lines: MAC-T, MACT-UV1, MACT-UV2. Cells were cultured for up to 12 d. Medium was collected during d 10, 11, and 12 in culture and subjected to SDS-PAGE followed by zymography in agarose-fibrin gels. A) Zymography was performed at standard conditions; B) zymography performed at standard conditions with the only difference being the addition of the u-PA specific inhibitor, amiloride (2 mM) in the agarose-fibrin gels; C)zymography performed at standard conditions with the only difference being the addition of .1 m g / d of anti-human tPA rabbit IgG in the agarose-fibrin gels. Lanes 1,2, and 3 were loaded with 50 pI of medium in which MAC-T, MACT-UV1, and MACTUV2 cells were cultured, respectively. The positions of the molecular weight markers are indicated to the left. t-PA = Tissue-PA, u-PA = urokinase-PA, t-PNPAI-I = complex of t-PA with PA inhibitor-1, u-PAPAI-1 = complex of u-PA with PA inhibitor-1.
Journal of Dairy Science Vol. 77, No. 10, 1994
HEEGARD ET AL
Figure 3. Zymographic analysis of plasminogen activator (PA) and activator-inhibitor (PAIPAI) complexes in the culture medium of two mammary epithelial cell lines: MACT-UVl, MACT-UV2. Cells were cultured for up to 12 d. Medium was collected during d 10, 11, and 12 in culture. Immunoprecipitation of proteins present in the culture medium was performed using anti-human PAL1 rabbit IgG as the first antibody and porcine anti-rabbit IgG as the secondary antibody. Culture medium after immunoprecipitation (50 pl) was subjected to SDS-PAGE, followed by zymography performed at standard conditions. Lane 1 was loaded with 10 pl of medium in which MACT-WI cells were cultured and was treated with the second antibody alone, for control purposes. Lanes 2 and 3 were loaded with 50 pl of medium in which MACT-Wl and MACT-UV2 were cultured and was treated with both antibodies. The positions of the molecular mass markers are indicated to the left. Several bands are present in lane 1, which was treated with the second antibody alone (no immunoprecipitation occurred), but the PNPAI complexes did not appear in lanes 2 and 3, which were treated with both antibodies (immunoprecipitation occurred).
areas appeared (Figure 3, lane 1). Figure 2B shows a very weak band (lanes 2 and 3) with an approximate molecular mass of 130,000 kDa in the medium of subclonal lines that could not be photographed (Figure 2A). This band remained unaffected in the presence of amiloride and never appeared following addition of anti-human t-PA IgG (Figure 2C) or following immunoprecipitation with antihuman PAI-1 IgG (Figure 3); this band probably represents a t-PARAI-1 complex. Heegard et al. (4, 5) unequivocally demonstrated the presence of a t-PAPAI-1 complex in bovine milk with a similar molecular mass of 130,000 kDa. When the culture medium was incubated in the presence of rabbit anti-human PAI-1 IgG, this treatment eliminated not only the activator inhibitor complexes, but also the 75,000-kDa tPA band (Figure 3). We have no explanation for this unusual observation, except that this polyclonal antibody may recognize an epitope in t-PA. Results of zymography apparently support an earlier conciusioi &ai subclonal iines- produce more PA than do the parental MAC-T cells (Table 2). First, subclonal lines produce tPA in addition to u-PA. Parental MAC-T cells produce only u-PA. Second, because equal amounts of protein were loaded in each lane, the intensity of the u-PA lysis areas apparently is much stronger in subclonal lines than in parental MAC-T cells (Figure 2A). U-PA Versus GPA
The most interesting finding of this study
was the exclusive production of u-PA by
parental MAC-T cells; the subclonal cells produced both types of PA (u-PA and t-PA) and PAI-1. Factors or conditions that govern the differences in the expression of the PA phenotype between parental and subclonal cell lines remain unclear. An earlier study (16) established that expression of PA may vary in fibrin, resulting in aFtive PA and inactive PAI- response to external stimuli, which does not 1 (2, 4, 5). This type of dissociation explains explain the differences observed in our study, why the complex still migrates at molecular because parental and subclonal lines were culmass of 120,000 kDa but is enzymatically tured under the same conditions. The major active to produce the lysis areas in Figure 2, A reason for the differences may be a general and C. For control, when the culture medium instability of the parental cells, following was incubated with only the second antibody prolonged subculturing. A previous study (20) (porcine anti-rabbit IgG), all of the main lysis established that MAC-T cells can be successJournal of Dairy Science Vol. 77, No. IO, 1994
PLASMINOGEN ACTIVATORS AND EPITHELIAL CELLS
fully subcloned, leading to generation of lines with different growth properties, cell size, chromosomal profile, and IGF-I receptor dynamics. Thus, not surprisingly, parental and newly generated subclonal lines differ in the amount and the type of PA produced. Significance of PA Produced by Mammary Epithelial Cells
UV1 and MACT-W2. The medium of the subclonal lines contained, in addition to u-PA, the 75,000-kDa t-PA and activator-inhibitor complexes. Further studies will examine whether known modulators of mammary growth and development affect PA and PA1 production by mammary epithelial cells and whether the enzyme can be used to monitor the physiological state of the gland (lactation vs. involution).
Several studies (4, 5 , 15, 19) have established that u-PA and t-PA are present in boACKNOWLEDGMENTS vine milk. Furthermore, both types of PA have been detected in mammary tissues (5, 14). The The research described herein was funded cellular origin of the PA is not known. Our by the Vermont Agricultural Experiment Stastudy suggests that PA and PA1 may be tion and the Walker Research Fund. produced, at least in part, by mammary epithelial cells. However, caution should be REFERENCES exercised because it is not known whether 1 Akers, R. M. 1990. Lactation physiology: a ruminant expression of PA is a common property of animal perspective. Protoplasma 159:96. mammary epithelial cells under select culture 2Andreasen, P. A., P. Kristensen, L. R. Lund, and K. conditions only or whether PA is also exDano. 1990. Urokinase-type plasminogen activator is pressed by mammary epithelial cells in vivo. increased in the involuting ventral prostate of casFurthermore, the MAC-T cells were produced trated rats. Endocrinology 126:2567. 3 Dano, K., P. A. Andreasen, J. Hansen-Grondahl, P. from primary bovine mammary epithelial cells Kristensen, and L. S. Nielsen. 1985. Plasminogen by stable transfection with a plasmid bearing activators, tissue degradation, and cancer. Adv. Canthe sequence for SV-40 large T-antigen (8). cer Res. 44:139. Dan0 et al. (3) reported that cultures of cells 4Heegard. C. W., T. Christensen, and P. A. Andreasen. 1993. t-PA binds to casein micelles in bovine milk. transformed by DNA viruses often produced Page 64 in Proc.Mol. Cell. Biol. Plasminogen Activamore PA than the nontransformed cultures tion. Cold Spring Harbor, NY. from which they were derived. Transformation 5 Heegard, C. W., T. Christensen, M. D. Rasmussen, C. procedures may select cells for higher PA Benfeldt, N. E. Jensen, K. Sejrsen, T. E. Petersen, and production. P. A. Andreasen. 1994. Plasminogen activators in bovine milk during mastitis, an inflammatory disease. In general, u-PA expression has been asFibrinolysis 8:22. sociated with extracellular proteolytic func6Hurley. W. L. 1986. Identification of plasminogen tions, such as tissue remodeling, inflammation, activator and protease activities in mammary secreor cell migration (16). Involution of the mamtions during the dry period. J. Dairy Sci. 69(Suppl. 1): mary gland is an example of tissue remodeling 205.(Abstr.) 7 Hurley, W. L. 1989. Mammary gland function during in the adult, in terms of the progression of involution. J. Dairy Sci. 72:1637. anatomical changes, and the gradual transition 8 Huynh, H. T., G. Robitaille, and J. D. Turner. 1991. of the gland from an active to inactive and Establishment of bovine mammary epithelial cells then return to an active state. Following sud(MAC-T): an in vitro model for bovine lactation. Exp. den cessation of milking, the events occurring Cell Res. 197:191. 9 Lu, D. D., and S. S. Nielsen. 1993. Isolation and in the mammary gland involve proteolysis. characterization of native bovine milk plasminogen Possibly, u-PA provides this proteolytic acactivators. J. Dairy Sci. 76:3369. tivity. 10Nielsen. L. S., J. G. Hansen, P. A. Andreasen, L. CONCLUSIONS
Production of PA was examined in three mammary epithelial cell lines. We detected a 50,000-kDa u-PA form in the medium of parental (MAC-T) and subclonal lines, MACT-
Skriver, K. Dano, and J. Zeuthen. 1983. Monoclonal antibody to human 66,000 molecular weight plasminogen activator from melanoma cells. Specific enzyme inhibition and one-step affinity purification. Eur. Mol. Biol. Organ. J. 2:115. 11 Oliver, S. P., and L. M. Sordillo. 1989. Approaches to manipulation of mammary involution. J. Dairy Sci. 72:1647. Journal of Dairy Science Vol. 77, No. 10, 1994
HEEGARD ET AL.
12Ossowski. L., D. Biegel. and E. Reich. 1979. Mammary plasminogen activator: correlation with involution, hormonal modulation and comparison between normal and neoplastic tissue. Cell 6:929. 13 Politis, I.. D. M. Barbano, and R. C. Gorewit. 1992. Changes in plasminogen activator in different physiological states of the gland. J. Dairy Sci. 75 (Suppl. 1): 181.(Ab*.) 14 Politis. I., B. Zavizion, D. M. Barbano. J. D. Turner, and R. C. Gorewit. 1992. Plasminogen activator activity in cultured mammary epithelial cells. J. Anim. Sci. (Suppl. 1):213.(Abstr.) 15 Politis. I., X.Zhao, B. W. McBride, and J. H. Burton. 1991. Plasminogen activator production by blood monocytes and milk macrophages. Am. J. Vet. Res. 52:1208. 16 Saksela, 0..and D. B. Rifkin. 1988. Cell-associated plasminogen activation: regulation and physiological functions. Annu. Rev. Cell Biol. 4:93.
Journal of Dairy Science Vol. 77, No. 10, 1994
17Turner, J. D., and H. T. Huynh. 1991. Role of tissue remodelling in mammary epithelial cell proliferation and morphogenesis. J. Dairy Sci. 74:2801. 18 Vassali, J.-D., and D. Belin. 1987. Amiloride selectively inhibits the urokinase-type plasminogen activator. Fed. Exp. Biol. Soc. Lett. 214:187. 19Zachos, T., I. Politis, R. C. Gorewit. and D. M. Barbano. 1992. Effect of mastitis on plasminogen activator activity of milk somatic cells. J. Dairy Res. 59:461. 2 0 Z a m m n . C., H. R. Lijnen, and D. Collen. 1984. Kinetics of the activation of plasminogen by natural and recombinant tissue-type plasminogen activator. J. Biol. Chem. 259:2080. 21 Zavizion, B., I. Politis, and R. C. Gorewit. 1993. Subcloning of the MAC-T clonal mammary epithelial cell line 1. Cell morphology, growth properties, and casein production. J. Dairy Sci. 7qSuppl. 1): 171.(Abstr.)