Bovine Mammary Myoepithelial Cells. 1. Isolation, Culture, and Characterization B. ZAVIZlON, I. POUTIS, and R. C. GOREWIT Lactation Physiology Laboratory Department of Animal Science Cornell University Ithaca, NY 14853 ABSTRACT
The objective of this study was to isolate, purify, culture, and characterize myoepithelial cells from bovine mammary glands. Myoepithelial cells were separated from other mammary and blood cells after collagenase digestion and centrifugation using metrizoate-ficoll gradients. Myoepithelial cells were identified by their characteristic morphology and cloned using selective detachment. They contained many densely packed myofilaments, very few cytoplasmic organelles, elongated surface projections, and a dense, irregularly shaped nuclei. Some cells were as large as 1.2 mm in culture. Myoepithelial cells contained an extensive network of cytoskeletal proteins, including a-smooth muscle actin, a-actinin, and vimentin. When cultured, they tended to repel one another and never grew as closely associated cells. The myoepithelial nature of these cells was verified by showing that they contracted in response to oxytocin, bound oxytocin, and did not produce casein. Myoepithelial cells from fetal and lactating glands grew very well in culture. Active division of myoepithelial cells could be maintained for at least 3 mo, and cells could be serially subcultured at least seven times. The successful isolation and culture of bovine mammary myoepithelial cells make utilization of these cells possible in order to study their role in mammary growth and differentiation and milk ejection. (Key words: bovine myoepithelial cells, epithelial cells)
Received May 13,
1m.
Accepted July 20, 1992. 1992 J Dairy Sci 75:3367-3380
Abbreviation key: BSA = bovine serum albumin, FITC = fluorescein isothiocyanate isomer I, HBSS = Hanks balanced salt solution, TRITC = tetramethylrhodamine isothiocyanate. INTRODUCTION
The myoepithelium surrounds the epithelium of alveoli and mammary ducts forming a basket weave of branching satellite cells around alveoli but does not form a true syncytium (9, 16). Myoepithelial cells ultrastructurally resemble smooth muscle in their content of filamentous tracts, which fill most of the cytoplasmic volume in differentiated cells. Myoepithelial cells in situ contract in response to oxytocic compounds, which results in compression of the alveolus and milk flow from the lumen into the ductal areas of the gland (9, 16). In vitro studies on the physiology and biochemistry of myoepithelial cells have been hampered by the difficulty of purifying and growing these cells in long-term monolayer cultures. Soloff et a1. (15) were able to separate rat mammary myoepithelial from epithelial cells because myoepithelial cells remained connected to one another after collagenase digestion. However, all morphological and functional characteristics of rat myoepithelial cells were determined immediately after purification. No attempts were made to grow these cells in culture (15). Success in extending the growth of myoepithelial cells in vitro would greatly facilitate investigations of the role that these cells play in mammary growth and differentiation and milk ejection. The objective of our experiments was to establish primary cultures of bovine mammary myoepithelial cells and to determine their properties and growth characteristics in vitro. Morphology, cellular physiol3367
3368
ZAVIZION ET AL.
ogy, and growth characteristics of primary mammary epithelial cells were investigated for comparative purposes. MATERIALS AND METHODS
Cows
Mammary tissue was obtained from cows in the following physiological states: 1) lactating. not pregnant; 2) not lactating. not pregnant; and 3) not lactating, pregnant at slaughter. Additionally, fetal mammary tissue was obtained from a heifer at 220 d of gestation and from a 4-yr-old, pregnant, nonlactating Holstein cow. Preparation of Primary Bovine Mammary Cells
Immediately after slaughter of cows, pieces of mammary tissue (10 to 30 g) were excised, placed in medium 199 (Sigma Chemical Co., S1. Louis, MO) containing penicillin (50 IV/ ml) and streptomycin (50 IJ.gtml), and stored on ice (O°C) until dissociation. Tissue was cut into slices (l to 3 mm 3 thick), washed in PBS, pH 7.2, and incubated in medium 199 containing 1.6 mgtml of collagenase type IV (I g of tissuel5 ml of medium; Sigma Chemical Co.). Tissue was stirred vigorously with a magnetic bar. Following incubation (6 to 10 h) at room temperature (24°C), the suspension was centrifuged at 150 x g for 15 min at 4°C. The cell pellet, containing undigested tissue, was resuspended in medium 199 and filtered through three layers of cheesecloth. The filtrate was centrifuged at 150 x g for 15 min and washed twice in PBS containing penicillin (50 IV/ml) and streptomycin (50 IJ.glml) before it was subjected to gradient centrifugation. Cell pellets were resuspended in 5 ml of medium 199 containing DNase I (100 IJ.glml) and layered on a preformed gradient Lymphoprep® (Nycomed, Oslo, Norway). Briefly, 5 ml of medium were layered on top of 10 ml of Lymphoprep90% viable), and processed further as described later. Cell Culture Procedures
Mammary cells were grown in plastic petri dishes (Coming Glass Works, Coming, NY). Some dishes were treated with gelatin (.1 mgt ml) or 7.5 IJ.glml of mouse type N collagen (Sigma Chemical Co.). Cultures were maintained at 37"C in a humidified C02 incubator. Medium consisted of a 1:1 (vol/vol) mixture of RPMI-I640 (Sigma Chemical Co.). and Iscove's Modified Dulbecco's Modified Eagle Medium (Sigma Chemical Co.) supplemented with 10% horse serum, 5% fetal calf serum (Sigma Chemical Co.), penicillin (100 IV/ml), and streptomycin (100 IJ.glml). Medium was changed every day during wk 1 of culture, three times a week thereafter, and every 24 h for experimental cultures. Secretory epithelial and myoepithelial cells were cloned from primary colonies following addition of versene (.02% EDTA; Sigma Chemical Co.) in Hanks balanced salt solution (BBSS; Sigma Chemical Co.) without calcium and magnesium, followed by addition of versene plus .01 % trypsin in HBSS until the desired percentage of cells had detached. Cells from colonies with similar morphology were manually harvested using a micromanipulator (Leitz, Bremme, Germany) fitted with a glass pipette. Cells were visualized with a Nikon (Nippon Kogaku K.K., Tokyo, Japan) inverted microscope. Homogenecity of cultures was verified using morphological characteristics and additional cloning from colonies. Growth curves were obtained after seeding 1000 to 5000 cells per well in 24-well multiwells (Becton Dickinson, Lincoln Park, NJ). Cells were grown on plastic and incubated for 6 to 10 d. Cells were removed by complete trypsinization (10 to 30 min at 37°C) every 24 h. Total cells were counted using a hemocytometer.
MYOEPITHELIAL CELLS
Scanning Electron Microscopy
Scanning electron microscopy was performed on isolated cells using the method of Cohen (2). Specimens were observed using an Amray 1000 scanning electron microscope (Amray Inc., Bedford, MA). Transmission electron microscopy was performed on isolated cel1s according to the procedure of Soloff et al. (15). Sections were observed using a Phillips 300 electron microscope (Phi11ips Electronics, Mahwah, NJ). Response to Oxytocin
In order to carry out dose-response chal1enges with oxytocin, myoepithelial and epithelial cel1s were seeded onto glass cover slips within 35-mm culture dishes for approximately 3 to 5 d. During this time, cells formed semiconfluent monolayers. Cover slips containing cells were washed gently with warm (37°C) medium 199, washed with HBSS, and then placed in 35-mm petri dishes. One milliliter of oxytocin (Calbiochem, La Jolla, CA) 00-9 to 10-5 M) dissolved in HBSS containing .9 mM magnesium and 1.2 mM calcium was added, and coverslips were incubated at 37°C in a humidified C~ incubator for 20 min. Changes in cell morphology were observed using an inverted microscope (Nikon TMS; Nippon Kogaku K.K.) with phase contrast optics. Immunofluorescence of Cytoskeletal Proteins
To characterize cytoskeletal elements, myoepithelial cells were cultured on glass coverslips for approximately 3 to 5 d. During this time, cells formed semiconfluent monolayers. Cells were then washed with warm (37°C) PBS and fixed in either glutaraldehyde (1.5%) for 30 min or cold (-40°C) acetone and methanol (1:1, voVvol) for 30 s. Cells were rinsed several times with PBS and incubated with a series of primary monoclonal antibodies: mouse anti-a-actinin (Sigma Chemical Co.), mouse anti-a smooth muscle actin (Sigma Chemical Co.), and mouse anti-vimentin (Boehringer Mannheim, Indianapolis, IN). All antibodies were used according to manufacturers' recommendations and diluted in PBS plus .5% BSA and .1 % sodium azide. Cover-
3369
slips were incubated with primary antibodies at room temperature (24°C) in a humidified chamber for 1 to 2 h. Specificity of staining was verified by omitting the addition of the primary antibody or by incubation with nonimmune serum. Coverslips were subjected to three 5-min washes in PBS, PBS plus .5% NP40 (Sigma Chemical Co.), and PBS plus .5% NP-4O plus 1.5% BSA. Following the last wash, cells were incubated with secondary antibody (anti-mouse IgG-fluorescein isothiocyanate isomer I (FITC) conjugate or antimouse IgG-tetramethylrhodamine isothiocyanate (TRITC) conjugate (Sigma Chemical Co.). All secondary antibodies were used in working dilutions (PBS plus .5% BSA plus .1 % sodium azide) according to manufacturers' recommendations. Coverslips were incubated with secondary antibody at room temperature (24°C) in a humidified chamber for 30 to 60 min and washed as described using the same set of solutions but in reverse order. For oxytocin-binding experiments, cells were treated with oxytocin (10-7 to 10-5 M) as described. The binding of oxytocin to myoepithelial cells was confirmed by indirect immunofluorescence using affinity-purified rabbit anti-oxytocin (Calbiochem) as the primary antibody, added at a dilution of 1:16. The second antibody was goat anti-rabbit FITC conjugate, used in dilutions according to manufacturers' recommendations. Coverslips were placed on microscope slides and visualized with a Nikon Optiphot fluorescence microscope, equipped with epifluorescence optics and appropriate excitatory and barrier filters for FITC and TRITC. Casein Analysis
Western blot analysis using monoclonal antibodies recognizing as-casein and ~-casein was employed to examine the presence or absence of as-casein and ~-casein in secretory epithelial and myoepithelial cell extracts. Two established mammary epithelial cell lines were used to examine casein production by epithelial cells: Mac-T (6) and BME (5), which were generously donated by J. D. Turner (McGill University, Montreal, PQ, Canada) and C. R. Baumrucker (The Pennsylvania State University, State College), respectively. Both antibodies were donated by C. W. Beatty Journal of Dairy Science Vol. 75, No. 12, 1992
3370
ZAVIZION ET AL.
(University of Chicago, Chicago, IL), and all details regarding their specificity were described earlier (8). One milliliter of PBS containing approximately 5 x 1()6 viable epithelial or myoepithelial cells grown on collagen (6) were sonicated (Branson, Danbury, for 45 s. The sonicate was filtered through a modified cellulose cell debris remover (Whatman Biosystems, Maidstone, England) and stored at -20·C until used. Then, SDS-PAGE was carried out using 12% acrylamide-resolving gels and 4% acrylamide-stacking gels as described by Politis et al. (13). Gels containing fractionated proteins present in cell extracts of secretory epithelial and myoepithelial cells were further processed for Western blots using methods of Politis et al. (13). The only difference was that nitrocellulose membranes were washed and incubated for 12 h with anti-ascasein or anti-~-casein IgG in ascites fluid added at a dilution of I:500. Membranes were then washed and incubated for 2 h with secondary antibody, peroxidase-conjugated antimouse IgG (1:1000 dilution; Sigma Chemical Co.). Membranes were washed and stained with .003% hydrogen peroxide and 4-chloroI-naphthol (.6 mg/ml; Bio-Rad Lab, Melville,
en
NY). Statistical AnalysIs
Differences in mean doubling times between fetal and adult myoepithelial cells were evaluated using Student's t test (P < .05). RESULTS AND DISCUSSION General Morphology
Electron microscopic studies revealed that individual isolated secretory epithelial and myoepithelial cells had distinct morphological differences. Isolated myoepithelial cells were larger than epithelial cells and had an involuted membrane (Figure IA). Isolated epithelial cells had a fairly uniform circular shape (Figure lB). Myoepithelial cells contained many densely packed myofilaments, very few cytoplasmic organelles, elongated surface projections, and extensive cytoplasmic Journal of Dairy Science Vol. 75. No. 12. 1992
membrane (Figure 2A). The nucleus was extremely dense and had an irregular shape (Figure 2A). These findings are similar to those of Soloff et al. (15) for myoepithelial cell isolates obtained from lactating rats. Secretory cells contained an extensive smooth and rough endoplasmic reticulum, ribosomes, Golgi apparati, and clusters of casein micelles (Figure 2B). Seven distinct preparations of primary cells (four myoepithelial and three epithelial cells) were utilized for morphological and growth characteristics. Myoepithelial cells were isolated from fetal; nonpregnant, nonlactating; pregnant, nonlactating; or nonpregnant, lactating cows. Myoepithelial cells from all sources had very similar size (80 to 160 JlIll) and morphology, never grew in a typical pattern, and tended to repel each other (Figure 3A). At confluence, they maintained the atypical pattern (Figure 3B), and they did not resemble the characteristic cobblestone morphology of epithelial cells or the fusiform morphology of fibroblasts. However, epithelial cells cultured on plastic displayed characteristic cuboidal or polygonal epithelial-like morphology (5, 6) and remained flattened (Figure 3C). When plated at low densities, epithelial cells quickly formed colonies with compact cobblestone morphology, showing the typical epithelial pavement of closely associated cells (Figure 3C). When primary cultures were allowed to remain confluent for several weeks in growth medium, epithelial cells rarely formed multilayers. The morphology of the majority (85 to 90%) of epithelial cells from nonpregnant, nonlactating; pregnant, nonlactating; and nonpregnant, lactating cows was similar. The major diversity of size and shape in the remaining 10 to 15% of the epithelial cell populations was observed in cells obtained from glands of nonpregnant, lactating cows at initial passage. The epithelial cells in that small portion were large (250 to 350 J.1D1) and had a very limited number of dividing cells, which were lost after the first passage. Growth Properties
In vitro growth properties of cloned mammary cells were examined. Following subcul-
MYOEPITHELIAL CELLS
3371
Figure 1. Scanning electron micrographs of bovine mammary myoepithelial (A) and epithelial cells (B). The myoepithelial cell has involuted membranes, but the epithelial cell has a fairly unifonn circular shape.
Journal of Dairy Science Vol. 75, No. 12, 1992
3372
ZAVIZION ET AL.
Figure 2. Transmission electron micrographs of bovine manunary myoepithelial (A) and epithelial (B) cells. The myoepithelial cell has an irregularly shaped nucleus (N), very few cytoplasmic organelles. elongated surface projections, and extensive cytoplasmic membrane. Epithelial cell contains an extensive smooth and rough endoplasmic reticulum (ER). Golgi apparatus. and milk caseins (me).
Journal of Dairy Science Vol. 75, No. 12, 1992
MYOEPITHELIAL CELLS
3373
Figure 3. Phase cootrast photomicrographs of preconfluent (A) and confluent (B) primary cultures of bovine manunary myoepithelial cells. A preconfluent colony of primary epithelial cells (e) is shown for comparative purposes. The characteristic cobblestone morphology of epithelial cells is present; myoepithelial cells showed minimal contact between them, even in confluent monolayers. x130. Journal of Dairy Science Vol. 75. No. 12, 1992
3374
ZAVIZION ET AL.
ture, myoepithelial cells progressed through a characteristic lag phase within the fIrst 24 h and entered into a log phase during the next 24 h in culture. Cell number increased for the next 4 to 5 d. Myoepithelial cells reached a cell density of 2 to 3 x 1()4/cm2 durin~ the ftrst or second passages and .5 x 1()4/cm during the sixth passage (Figure 4, A and B). In parallel, doubling times of fetal myoepithelial cells, calculated from growth curves, were approximately 22 and 40 h during the second and sixth passages, respectively (Figure 4A). Doubling times of myoepithelial cells obtained from lactating cows were 15 and 24 h during the second and sixth passages, respectively. Differences in doubting times between fetal and adult myoepithelial cells were statistically signifIcant (P < .05). During the ftrst to third passage, myoepithelial cells showed colony-forming ability (3 to 7%) at very low cell densities. No colony-forming ability was observed for myoepithelial cells of later than fifth passage. The inability of myoepithelial cells at later passages to form colonies is not an unusual phenomenon. All primary cells have a finite ability to proliferate and to form colonies. The progressive decline in proliferative capacity and, therefore, in colony-forming ability is usually deftned as ftnite life span phenotype or cellular senescence (11). Myoepithelial cells from pregnant, nonlactating cows were passed three times (after cloning) before appearance of nondividing, senescent cells. Those cells were very large (up to 1.2 rom) (10 to 20 times larger than proliferating cells) and remained in culture for 90 d. A small number of myoepithelial cells gave rise to colonies containing up to 200 to 300 cells, but then all cells were unable to divide. Myoepithelial cells from fetal and lactating glands grew very well in culture for at least 3 mo (>7 passages; >15 doubling). All myoepithelial cells showed ftbroblast-like morphology and maintained that morphology before cloning (0 passage). During cloning and after one to two passages after cloning, myoepithelial cells that were homogenous at the beginning of culture diverged into mixed populations containing at least two different size cell types: small proliferating and medium-sized cells (Figure SA). The large senescent (nondividing) cells predominated in culture after six to seven passages (Figure 5B). Journal of Dairy Science Vol. 75, No. 12, 1992
A
-'
-' UJ
3: a:
UJ
a.
(/)
-' -' W
o
o
234 B
5
6
7
5
6
7
DAYS
..J ..J W
3:
a: w a.
(/)
-' -' W
o
o
234
DAYS
Figure 4. Growth curves for bovine mammary myoepithelial cells from fetal (A) and lactating (8) mammary glands. Growth curves were derived using cells of the second (a) and sixth passages (e). Population doubling times of fetal myoepithelial cells were 22 and 40 h during the second and sixth passages, respectively. Population doubling times of myoepithelial cells from lactating cows were 15 and 24 h during the second and sixth passages, respectively. Results are representative of three independent experiments.
MYOEPITHELIAL CELLS
3375
Figure 5. Phase contrast photomicrographs of bovine fetal mammary myoepithelial cells during the second (A) and sixth passages (B). Numerous small and medium proliferating myoepithelial cells (A) and large. senescent (nondividing) cells (B) are present. x130.
Journal of Dairy Science Vol. 75. No. 12. 1992
3376
ZAVIZION ET AL.
Figure 6. ImmunolocaIization of vimentin (A), a.-aetinin (B), and a.-smooth muscle actin (C) in bovine manunary myoepithelial cells cultured on plastic. Primary antibodies were mouse monoclonal antibodies against a.-smooth muscle actin. a.-actinin. and vimentin. Secondary antibody was anti-mouse-IgG-fluorescein isothiocyanase isomer I. The nonspecific binding of the second antibody was low (0). >
The objective of this study was to isolate, purify, culture, and characterize myoepithelial cells from bovine mammary glands. Myoepithelial cells were separated from other mammary and blood cells after collagenase digestion and centrifugation using metrizoate-ficoll gradients. Myoepithelial cells were identified by their characteristic morphology and cloned using selective detachment. They contained many densely packed myofilaments, very few cytoplasmic organelles, elongated surface projections, and a dense, irregularly shaped nuclei. Some cells were as large as 1.2 mm in culture. Myoepithelial cells contained an extensive network of cytoskeletal proteins, including a-smooth muscle actin, a-actinin, and vimentin. When cultured, they tended to repel one another and never grew as closely associated cells. The myoepithelial nature of these cells was verified by showing that they contracted in response to oxytocin, bound oxytocin, and did not produce casein. Myoepithelial cells from fetal and lactating glands grew very well in culture. Active division of myoepithelial cells could be maintained for at least 3 mo, and cells could be serially subcultured at least seven times. The successful isolation and culture of bovine mammary myoepithelial cells make utilization of these cells possible in order to study their role in mammary growth and differentiation and milk ejection. (Key words: bovine myoepithelial cells, epithelial cells)
Received May 13,
1m.
Accepted July 20, 1992. 1992 J Dairy Sci 75:3367-3380
Abbreviation key: BSA = bovine serum albumin, FITC = fluorescein isothiocyanate isomer I, HBSS = Hanks balanced salt solution, TRITC = tetramethylrhodamine isothiocyanate. INTRODUCTION
The myoepithelium surrounds the epithelium of alveoli and mammary ducts forming a basket weave of branching satellite cells around alveoli but does not form a true syncytium (9, 16). Myoepithelial cells ultrastructurally resemble smooth muscle in their content of filamentous tracts, which fill most of the cytoplasmic volume in differentiated cells. Myoepithelial cells in situ contract in response to oxytocic compounds, which results in compression of the alveolus and milk flow from the lumen into the ductal areas of the gland (9, 16). In vitro studies on the physiology and biochemistry of myoepithelial cells have been hampered by the difficulty of purifying and growing these cells in long-term monolayer cultures. Soloff et a1. (15) were able to separate rat mammary myoepithelial from epithelial cells because myoepithelial cells remained connected to one another after collagenase digestion. However, all morphological and functional characteristics of rat myoepithelial cells were determined immediately after purification. No attempts were made to grow these cells in culture (15). Success in extending the growth of myoepithelial cells in vitro would greatly facilitate investigations of the role that these cells play in mammary growth and differentiation and milk ejection. The objective of our experiments was to establish primary cultures of bovine mammary myoepithelial cells and to determine their properties and growth characteristics in vitro. Morphology, cellular physiol3367
3368
ZAVIZION ET AL.
ogy, and growth characteristics of primary mammary epithelial cells were investigated for comparative purposes. MATERIALS AND METHODS
Cows
Mammary tissue was obtained from cows in the following physiological states: 1) lactating. not pregnant; 2) not lactating. not pregnant; and 3) not lactating, pregnant at slaughter. Additionally, fetal mammary tissue was obtained from a heifer at 220 d of gestation and from a 4-yr-old, pregnant, nonlactating Holstein cow. Preparation of Primary Bovine Mammary Cells
Immediately after slaughter of cows, pieces of mammary tissue (10 to 30 g) were excised, placed in medium 199 (Sigma Chemical Co., S1. Louis, MO) containing penicillin (50 IV/ ml) and streptomycin (50 IJ.gtml), and stored on ice (O°C) until dissociation. Tissue was cut into slices (l to 3 mm 3 thick), washed in PBS, pH 7.2, and incubated in medium 199 containing 1.6 mgtml of collagenase type IV (I g of tissuel5 ml of medium; Sigma Chemical Co.). Tissue was stirred vigorously with a magnetic bar. Following incubation (6 to 10 h) at room temperature (24°C), the suspension was centrifuged at 150 x g for 15 min at 4°C. The cell pellet, containing undigested tissue, was resuspended in medium 199 and filtered through three layers of cheesecloth. The filtrate was centrifuged at 150 x g for 15 min and washed twice in PBS containing penicillin (50 IV/ml) and streptomycin (50 IJ.glml) before it was subjected to gradient centrifugation. Cell pellets were resuspended in 5 ml of medium 199 containing DNase I (100 IJ.glml) and layered on a preformed gradient Lymphoprep® (Nycomed, Oslo, Norway). Briefly, 5 ml of medium were layered on top of 10 ml of Lymphoprep90% viable), and processed further as described later. Cell Culture Procedures
Mammary cells were grown in plastic petri dishes (Coming Glass Works, Coming, NY). Some dishes were treated with gelatin (.1 mgt ml) or 7.5 IJ.glml of mouse type N collagen (Sigma Chemical Co.). Cultures were maintained at 37"C in a humidified C02 incubator. Medium consisted of a 1:1 (vol/vol) mixture of RPMI-I640 (Sigma Chemical Co.). and Iscove's Modified Dulbecco's Modified Eagle Medium (Sigma Chemical Co.) supplemented with 10% horse serum, 5% fetal calf serum (Sigma Chemical Co.), penicillin (100 IV/ml), and streptomycin (100 IJ.glml). Medium was changed every day during wk 1 of culture, three times a week thereafter, and every 24 h for experimental cultures. Secretory epithelial and myoepithelial cells were cloned from primary colonies following addition of versene (.02% EDTA; Sigma Chemical Co.) in Hanks balanced salt solution (BBSS; Sigma Chemical Co.) without calcium and magnesium, followed by addition of versene plus .01 % trypsin in HBSS until the desired percentage of cells had detached. Cells from colonies with similar morphology were manually harvested using a micromanipulator (Leitz, Bremme, Germany) fitted with a glass pipette. Cells were visualized with a Nikon (Nippon Kogaku K.K., Tokyo, Japan) inverted microscope. Homogenecity of cultures was verified using morphological characteristics and additional cloning from colonies. Growth curves were obtained after seeding 1000 to 5000 cells per well in 24-well multiwells (Becton Dickinson, Lincoln Park, NJ). Cells were grown on plastic and incubated for 6 to 10 d. Cells were removed by complete trypsinization (10 to 30 min at 37°C) every 24 h. Total cells were counted using a hemocytometer.
MYOEPITHELIAL CELLS
Scanning Electron Microscopy
Scanning electron microscopy was performed on isolated cells using the method of Cohen (2). Specimens were observed using an Amray 1000 scanning electron microscope (Amray Inc., Bedford, MA). Transmission electron microscopy was performed on isolated cel1s according to the procedure of Soloff et al. (15). Sections were observed using a Phillips 300 electron microscope (Phi11ips Electronics, Mahwah, NJ). Response to Oxytocin
In order to carry out dose-response chal1enges with oxytocin, myoepithelial and epithelial cel1s were seeded onto glass cover slips within 35-mm culture dishes for approximately 3 to 5 d. During this time, cells formed semiconfluent monolayers. Cover slips containing cells were washed gently with warm (37°C) medium 199, washed with HBSS, and then placed in 35-mm petri dishes. One milliliter of oxytocin (Calbiochem, La Jolla, CA) 00-9 to 10-5 M) dissolved in HBSS containing .9 mM magnesium and 1.2 mM calcium was added, and coverslips were incubated at 37°C in a humidified C~ incubator for 20 min. Changes in cell morphology were observed using an inverted microscope (Nikon TMS; Nippon Kogaku K.K.) with phase contrast optics. Immunofluorescence of Cytoskeletal Proteins
To characterize cytoskeletal elements, myoepithelial cells were cultured on glass coverslips for approximately 3 to 5 d. During this time, cells formed semiconfluent monolayers. Cells were then washed with warm (37°C) PBS and fixed in either glutaraldehyde (1.5%) for 30 min or cold (-40°C) acetone and methanol (1:1, voVvol) for 30 s. Cells were rinsed several times with PBS and incubated with a series of primary monoclonal antibodies: mouse anti-a-actinin (Sigma Chemical Co.), mouse anti-a smooth muscle actin (Sigma Chemical Co.), and mouse anti-vimentin (Boehringer Mannheim, Indianapolis, IN). All antibodies were used according to manufacturers' recommendations and diluted in PBS plus .5% BSA and .1 % sodium azide. Cover-
3369
slips were incubated with primary antibodies at room temperature (24°C) in a humidified chamber for 1 to 2 h. Specificity of staining was verified by omitting the addition of the primary antibody or by incubation with nonimmune serum. Coverslips were subjected to three 5-min washes in PBS, PBS plus .5% NP40 (Sigma Chemical Co.), and PBS plus .5% NP-4O plus 1.5% BSA. Following the last wash, cells were incubated with secondary antibody (anti-mouse IgG-fluorescein isothiocyanate isomer I (FITC) conjugate or antimouse IgG-tetramethylrhodamine isothiocyanate (TRITC) conjugate (Sigma Chemical Co.). All secondary antibodies were used in working dilutions (PBS plus .5% BSA plus .1 % sodium azide) according to manufacturers' recommendations. Coverslips were incubated with secondary antibody at room temperature (24°C) in a humidified chamber for 30 to 60 min and washed as described using the same set of solutions but in reverse order. For oxytocin-binding experiments, cells were treated with oxytocin (10-7 to 10-5 M) as described. The binding of oxytocin to myoepithelial cells was confirmed by indirect immunofluorescence using affinity-purified rabbit anti-oxytocin (Calbiochem) as the primary antibody, added at a dilution of 1:16. The second antibody was goat anti-rabbit FITC conjugate, used in dilutions according to manufacturers' recommendations. Coverslips were placed on microscope slides and visualized with a Nikon Optiphot fluorescence microscope, equipped with epifluorescence optics and appropriate excitatory and barrier filters for FITC and TRITC. Casein Analysis
Western blot analysis using monoclonal antibodies recognizing as-casein and ~-casein was employed to examine the presence or absence of as-casein and ~-casein in secretory epithelial and myoepithelial cell extracts. Two established mammary epithelial cell lines were used to examine casein production by epithelial cells: Mac-T (6) and BME (5), which were generously donated by J. D. Turner (McGill University, Montreal, PQ, Canada) and C. R. Baumrucker (The Pennsylvania State University, State College), respectively. Both antibodies were donated by C. W. Beatty Journal of Dairy Science Vol. 75, No. 12, 1992
3370
ZAVIZION ET AL.
(University of Chicago, Chicago, IL), and all details regarding their specificity were described earlier (8). One milliliter of PBS containing approximately 5 x 1()6 viable epithelial or myoepithelial cells grown on collagen (6) were sonicated (Branson, Danbury, for 45 s. The sonicate was filtered through a modified cellulose cell debris remover (Whatman Biosystems, Maidstone, England) and stored at -20·C until used. Then, SDS-PAGE was carried out using 12% acrylamide-resolving gels and 4% acrylamide-stacking gels as described by Politis et al. (13). Gels containing fractionated proteins present in cell extracts of secretory epithelial and myoepithelial cells were further processed for Western blots using methods of Politis et al. (13). The only difference was that nitrocellulose membranes were washed and incubated for 12 h with anti-ascasein or anti-~-casein IgG in ascites fluid added at a dilution of I:500. Membranes were then washed and incubated for 2 h with secondary antibody, peroxidase-conjugated antimouse IgG (1:1000 dilution; Sigma Chemical Co.). Membranes were washed and stained with .003% hydrogen peroxide and 4-chloroI-naphthol (.6 mg/ml; Bio-Rad Lab, Melville,
en
NY). Statistical AnalysIs
Differences in mean doubling times between fetal and adult myoepithelial cells were evaluated using Student's t test (P < .05). RESULTS AND DISCUSSION General Morphology
Electron microscopic studies revealed that individual isolated secretory epithelial and myoepithelial cells had distinct morphological differences. Isolated myoepithelial cells were larger than epithelial cells and had an involuted membrane (Figure IA). Isolated epithelial cells had a fairly uniform circular shape (Figure lB). Myoepithelial cells contained many densely packed myofilaments, very few cytoplasmic organelles, elongated surface projections, and extensive cytoplasmic Journal of Dairy Science Vol. 75. No. 12. 1992
membrane (Figure 2A). The nucleus was extremely dense and had an irregular shape (Figure 2A). These findings are similar to those of Soloff et al. (15) for myoepithelial cell isolates obtained from lactating rats. Secretory cells contained an extensive smooth and rough endoplasmic reticulum, ribosomes, Golgi apparati, and clusters of casein micelles (Figure 2B). Seven distinct preparations of primary cells (four myoepithelial and three epithelial cells) were utilized for morphological and growth characteristics. Myoepithelial cells were isolated from fetal; nonpregnant, nonlactating; pregnant, nonlactating; or nonpregnant, lactating cows. Myoepithelial cells from all sources had very similar size (80 to 160 JlIll) and morphology, never grew in a typical pattern, and tended to repel each other (Figure 3A). At confluence, they maintained the atypical pattern (Figure 3B), and they did not resemble the characteristic cobblestone morphology of epithelial cells or the fusiform morphology of fibroblasts. However, epithelial cells cultured on plastic displayed characteristic cuboidal or polygonal epithelial-like morphology (5, 6) and remained flattened (Figure 3C). When plated at low densities, epithelial cells quickly formed colonies with compact cobblestone morphology, showing the typical epithelial pavement of closely associated cells (Figure 3C). When primary cultures were allowed to remain confluent for several weeks in growth medium, epithelial cells rarely formed multilayers. The morphology of the majority (85 to 90%) of epithelial cells from nonpregnant, nonlactating; pregnant, nonlactating; and nonpregnant, lactating cows was similar. The major diversity of size and shape in the remaining 10 to 15% of the epithelial cell populations was observed in cells obtained from glands of nonpregnant, lactating cows at initial passage. The epithelial cells in that small portion were large (250 to 350 J.1D1) and had a very limited number of dividing cells, which were lost after the first passage. Growth Properties
In vitro growth properties of cloned mammary cells were examined. Following subcul-
MYOEPITHELIAL CELLS
3371
Figure 1. Scanning electron micrographs of bovine mammary myoepithelial (A) and epithelial cells (B). The myoepithelial cell has involuted membranes, but the epithelial cell has a fairly unifonn circular shape.
Journal of Dairy Science Vol. 75, No. 12, 1992
3372
ZAVIZION ET AL.
Figure 2. Transmission electron micrographs of bovine manunary myoepithelial (A) and epithelial (B) cells. The myoepithelial cell has an irregularly shaped nucleus (N), very few cytoplasmic organelles. elongated surface projections, and extensive cytoplasmic membrane. Epithelial cell contains an extensive smooth and rough endoplasmic reticulum (ER). Golgi apparatus. and milk caseins (me).
Journal of Dairy Science Vol. 75, No. 12, 1992
MYOEPITHELIAL CELLS
3373
Figure 3. Phase cootrast photomicrographs of preconfluent (A) and confluent (B) primary cultures of bovine manunary myoepithelial cells. A preconfluent colony of primary epithelial cells (e) is shown for comparative purposes. The characteristic cobblestone morphology of epithelial cells is present; myoepithelial cells showed minimal contact between them, even in confluent monolayers. x130. Journal of Dairy Science Vol. 75. No. 12, 1992
3374
ZAVIZION ET AL.
ture, myoepithelial cells progressed through a characteristic lag phase within the fIrst 24 h and entered into a log phase during the next 24 h in culture. Cell number increased for the next 4 to 5 d. Myoepithelial cells reached a cell density of 2 to 3 x 1()4/cm2 durin~ the ftrst or second passages and .5 x 1()4/cm during the sixth passage (Figure 4, A and B). In parallel, doubling times of fetal myoepithelial cells, calculated from growth curves, were approximately 22 and 40 h during the second and sixth passages, respectively (Figure 4A). Doubling times of myoepithelial cells obtained from lactating cows were 15 and 24 h during the second and sixth passages, respectively. Differences in doubting times between fetal and adult myoepithelial cells were statistically signifIcant (P < .05). During the ftrst to third passage, myoepithelial cells showed colony-forming ability (3 to 7%) at very low cell densities. No colony-forming ability was observed for myoepithelial cells of later than fifth passage. The inability of myoepithelial cells at later passages to form colonies is not an unusual phenomenon. All primary cells have a finite ability to proliferate and to form colonies. The progressive decline in proliferative capacity and, therefore, in colony-forming ability is usually deftned as ftnite life span phenotype or cellular senescence (11). Myoepithelial cells from pregnant, nonlactating cows were passed three times (after cloning) before appearance of nondividing, senescent cells. Those cells were very large (up to 1.2 rom) (10 to 20 times larger than proliferating cells) and remained in culture for 90 d. A small number of myoepithelial cells gave rise to colonies containing up to 200 to 300 cells, but then all cells were unable to divide. Myoepithelial cells from fetal and lactating glands grew very well in culture for at least 3 mo (>7 passages; >15 doubling). All myoepithelial cells showed ftbroblast-like morphology and maintained that morphology before cloning (0 passage). During cloning and after one to two passages after cloning, myoepithelial cells that were homogenous at the beginning of culture diverged into mixed populations containing at least two different size cell types: small proliferating and medium-sized cells (Figure SA). The large senescent (nondividing) cells predominated in culture after six to seven passages (Figure 5B). Journal of Dairy Science Vol. 75, No. 12, 1992
A
-'
-' UJ
3: a:
UJ
a.
(/)
-' -' W
o
o
234 B
5
6
7
5
6
7
DAYS
..J ..J W
3:
a: w a.
(/)
-' -' W
o
o
234
DAYS
Figure 4. Growth curves for bovine mammary myoepithelial cells from fetal (A) and lactating (8) mammary glands. Growth curves were derived using cells of the second (a) and sixth passages (e). Population doubling times of fetal myoepithelial cells were 22 and 40 h during the second and sixth passages, respectively. Population doubling times of myoepithelial cells from lactating cows were 15 and 24 h during the second and sixth passages, respectively. Results are representative of three independent experiments.
MYOEPITHELIAL CELLS
3375
Figure 5. Phase contrast photomicrographs of bovine fetal mammary myoepithelial cells during the second (A) and sixth passages (B). Numerous small and medium proliferating myoepithelial cells (A) and large. senescent (nondividing) cells (B) are present. x130.
Journal of Dairy Science Vol. 75. No. 12. 1992
3376
ZAVIZION ET AL.
Figure 6. ImmunolocaIization of vimentin (A), a.-aetinin (B), and a.-smooth muscle actin (C) in bovine manunary myoepithelial cells cultured on plastic. Primary antibodies were mouse monoclonal antibodies against a.-smooth muscle actin. a.-actinin. and vimentin. Secondary antibody was anti-mouse-IgG-fluorescein isothiocyanase isomer I. The nonspecific binding of the second antibody was low (0). >
