Cell Proliferation and Milk Protein in Rabbit Mammary Cell Cultures Gene Expression

We analyzed the synthesis of DNA, the rate of cell proliferation, and the expression of milk protein genes in mammary cells grown as primary cultures on or in collagen gels in chemically defined media. We assessed DNA synthesis and cell growth, measured by [aH]thymidine incorporation into acid-insoluble material, DNA content, and cell counts, in a progesteroneand prolactin-containing medium. In some experiments, cultures were pulsed for I h with [3H]thymidine and dissociated into individual cells which were cytocentrifuged and processed for immunocytochemistry and autoradiography. We analyzed expression of milk protein genes at the transcriptional, translational, and posttranslational levels in a progesteronedepleted medium in the presence or absence of prolactin. We measured protein secretion by radioimmunoassays with antisera directed against caseins, ~-lactalbumin and milk transferrin. We determined protein synthesis by incorporating radio-labeled amino acids into acid-precipitable material and by immunoprecipitating biosynthetically labeled milk proteins. We assessed the accumulation of casein mRNA by hybridizing total cellular RNA extracted from cultured cells with 32p-labeled casein cDNA probes. On attached collagen gels, the cells synthesized DNA and replicated until they became confluent. The overall protein synthetic activity was low, and no milk proteins were synthesized or secreted even in the presence of prolactin. The block in milk protein gene expression was not restricted to translational or posttranslational events but also included transcription, since no casein mRNA accumulated in these cells. On floating gels, protein synthesis was threefold higher than in cells from attached gels. Overall protein synthesis as well as casein and ~-lactalbumin synthesis and secretion were prolactindependent with maximal stimulation at around 10 -9 M. A marked inhibition occurred at higher hormone concentrations. Casein mRNA accumulated in these cells, provided prolactin was present in the medium. In contrast, these cells did not synthesize DNA, nor did they replicate. In embedding gels, the rate of cell proliferation was exponential over 25 d with a doubling time of ~70 h. The overall protein synthesis increase was parallel in time with the increase in cell number. Caseins and ~-lactalbumin (in contrast to transferrin) were synthesized only in the presence of prolactin. We observed the same hormone dependency as with cells growing on floating gels. The number of caseinand transferrin-positive cells was measured after dissociating the cell cultures. At day 12, 60% of the total cells stored transferrin in small cytoplasmic vesicles, whereas only 25% of the cells accumulated casein. Differences in the organization and in the shape of mammary cells depending on cell surface conditions suggest that the geometry of the cells, their interaction with extracellular matrix constitutents, and cellto-cell interactions play a role in the expression of two mammary functions: DNA synthesis and growth, as well as milk protein gene expression. THE JOURNAL OF CELL BIOLOGY VOLUME 96 MAY 1983 1435-1442 © The Rockefeller University Press. 0021-9525/83/05/1435/08 $I.00 1 4 3 5 on D ecem er 3, 2017 jcb.rress.org D ow nladed fom The mechanisms regulating D N A synthesis and cell replication during pregnancy are poorly understood. It is known from in vitro studies that estrogens in conjunction with progesterone induce D N A synthesis in end-bud cells, whereas duct cells seem to respond preferentially to progesterone (3). Ovarian steroids, however, are not the only hormones involved in the stimulation of mammary gland growth. Adrenal steroids, insulinlike hormones, and thyroid and lactogenic hormones have been shown to induce lobulo-alveolar differentiation in rat and mouse (29, 43, 44). Their role has been examined in vitro by means of growing organ explants of mammary gland in chemically defmed media supplemented with insulin, glucoor mineralocorticoids, and prolactin used at pharmacological concentrations (18, 19). Since growth occurred in tbe absence of ovarian steroids, it has been speculated that endogenous levels of estrogen and progesterone carried over into the culture system might be sufficient for priming the tissue, thus rendering it independent from exogenous addition of these hormones. In the rabbit, there is a lack of information concerning tbe hormonal environment that initiates and sustains those physiological changes. The onset of milk protein gene expression and milk secretion at parturition and during lactation is induced and modulated by an interplay among a complex of hormones including steroid and polypeptide hormones that appear to act in a specific temporal sequence (36). The regulation of milk protein gene expression by lactogenic hormones has been extensively analyzed at the molecular level with complementary D N A hybridizing with casein and a-lactalbumin mRNAs as probes. Several species have been studied both in vivo and in vitro, using organ explants. It is now weU established in rabbit (9, 16, 17), rat (13, 33, 34, 36), mouse (1, 5, 40), and guinea pig (6) that lactogenic hormones, i.e., prolactin, placental lactogens, and primate growth hormones, regulate the gene expression of the caseins and of a-lactalbumin--a protein involved in lactose synthesis. In rat mammary organ cultures, prolactin exerts pleiotropic effects by increasing on the one hand casein transcription and on the other hand the half-life o f newly synthesized m R N A molecules (13). Progesterone inhibits both these transcriptional and posttranseriptional events (17, 35), whereas glucocorticoids, which are inactive alone, potentiate the effects of lactogenic hormones (8). There have been few studies reported on the regulation of milk protein gene expression at the transcriptional level either in primary mammary cell cultures or in established cell lines (41). Most studies using primary cultures of normal or transformed cells were restricted to translational or post-translational events. In such systems, the amount of casein and a-lactalbumin synthesized or secreted per cell remained far below the amounts produced in the intact lactating gland (12, 30, 31). Following the work of Pitelka and of her co-workers (11, 12, 45), we grew mammary cells in or on coUagenous matrices in serum-free chemically defined media, and we examined the role of cell shape and cytodifferentiation on the functional differentiation of mammary cells by morphological means (14). In the present paper, we analyze cell proliferation and D N A synthesis as well as milk protein gene expression at the transcriptional translational, and posttranslational levels under three different culture conditions, i.e., on attached, floating, or embedding collagen gels. We show that cell replication and milk protein gene expression depend both on a specific hormonal environment and on cell organization. Part of this work has been presented in an abstract form (23). 1436 T•E JOURNAL OF CELL BIOLOGY • VOLUME 96, 1983 MATERIALS AND METHODS Materials are given in the accompanying pap¢r (14). Cell Sampling: Pools of dispersed mammary ceils were routinely prepared from three individual rabbits, frozen, and stored in liquid nitrogen. To normalize our experimental data, prior to the exhaustion of one pool, we took a sample for comparison with the subs~uent pool. Culture Conditions: The standard medium for plating, growing, and biosynthetically labeling the cells was a 1 to 1 mixture of M199 and FI2 medium supplemented with gentamicin (100/~g/ml) and fungizone (2.5 /zg/ml) for the fast 5 d of culture. Cell aggregates (see MateriaLs and Methods of accompanying paper [14]), suspended in medium containing 20% horse serum and 5% fetal calf serum, were plated at a density of l0 ~ ceLLs/era ~ for cell proliferation and DNA synthesis studies, and at 5 x 105 cells~era 2 for milk protein gane expression analysis. The ceils were embedded at a density of 10S/m1 of collagen mixture for cell replication and at 10 e ceils/ml for milk protein gene expression experiments. The ceils were cultured for 24 h in the same serum containing medium and then in serum-free medium complemented with 0.25% bovine serum albumin, 10 -~° M dexamethasone, l0 -~° M 17-/~-estradioL 10 -s M triiodothyronine, 10 ng/nd epidermal growth factor, and prostaglandin F2a (PGF2a), and 5 gg/ml insulin. For sustained growth and DNA synthesis, the medium was supplemented witb 10 -s M progesterone and 5 x 10 -s urine prolactin (PRL) unless otherwise stated. For protein synthesis and secretion experiments, we removed progesterone and added PRL, at concentrations varying between 10 -s and 10 -~° M, to determine the dose-dependency. The medium was changed every day. Floating gels were obtained 2 or 5 d aRer plating by mechanically detaching the attached collagen gels with a Pasteur pipet. Protein Synthesis: The attached, floating, and embedding gels were digested for 2 h at 37°C, with purified collagenase (see MateriaLs and Methods of accompanying paper [ 14]). For immunoprecipitation studies, the cell layers or the outgrowths of six wells (35 mm in diameter) were pooled for each experimental determination, washed in methionine-free medium, and incubated with 2.5 ml of [~S]methionine (>1,200 Ci/mmol) at 200 pCi/ml. For assessment of overall protein synthesis by precipitation of acid-insoluble material, the cell layers and the outgrowths of two wells (16 mm in diameter) were pooled, washed in leucine-free medium, and incubated with 0.5 ml of [aH]leucine (>300 Ci/mmol) at 10 #Ci/ml. At tbe end of incubation, protease inbibitors were added at a final concentration of 1 mM for phenylmethylsulfonyl fluoride (PMSF), 5/~g/ml for pepstatin A, leupeptin, antipain, and aprotinin and 10 gg/ml for soybean trypsin inhibitor. The cells were centrifuged and the supematant was saved. 1-2 ml of HzO were added to the pellets, which were sonicated for 30 s at 4°C. Ahquots were used to measure the DNA content (21) and to determine the incorporation of radioactivity into acid-insoluble material, or for immunoprecipitation with anticaseins, anti-a-lactalbumin, and antitransferrin sera (7) according to Maccechini et at. (26). Samples of the medium (superuatant), the pellet, and the immunoprecipitates were analyzed by SDS PAGE (27), followed by fluorography (25). Protein Secretion: The medium from cells grown on attached or floating gels was collected each day. The gels were then digested, and the cells were recovered and lysed as described. The amount of milk protein (caseins a, ~¢, a-lactalbumin, and transferrin) was measured by radio-immunoassays using specific goat antisera (7), radio-iodinated purified milk proteins, and a sheep antirabbit F(ab')2 serum to precipitate the soluble immune complexes in a classical sandwich assay. The amount of secreted protein we estimated from standard curves estabLished with purified cold milk proteins (7) and the values were normalized to the DNA content and hence to the number of cells, assuming 7 pg DNA per cell (22), and expressed as nanograms of milk protein secreted per 24 h per 10 ~ cells. To determine protein accumulation in cells from embedding gels, the cells were dissociated with trypsin (2.5 gg/ml), cytocentrifuged, and labeled with anti-milk protein sera. DNA Synthesis and Cell Counts: The collagen gels were frst digested with collagenase for 1-3 h. The cells were then incubated at 37°C with 2 pCi/ml of [~H]thymidine for 1 h. The amount of 3H in TCA-insoluble material was counted in a liquid scintillation counter, and the values were normalized to DNA content. For cell counting, the cells were further treated with trypsin (2.5 gg/ml) and EDTA (1 mM) for 15 rain, dispersed with a siliconized Pasteur pipet, and counted in a hemocytometer. Some cells were cytocentrifuged and processed for autoradiography or immunocytochemistry as described below. Extraction of RNA and Blot Analysis: Total cellular RNA was extracted at 65°C with phenol (37) from ceils grown on attached, floating, or in embedding collagen gels. Total cellular RNA was denatured in 1 M glyoxal, 50% DMSO, 10 mM phosphate, pH 6.5, 0.1% SDS for 5 min at 50°C (following a modification of the procedure of McMaster and Carmichael [28]) and was clectrophoresed on a horizontal agarose slab gel in 10 mM phosphate buffer, pH 6.5. The RNA was transferred onto nitrocellulose filters following the protocol of on D ecem er 3, 2017 jcb.rress.org D ow nladed fom Thomas (42) and was hybridized to cloned 32p-labeled ,/-casein cDNA (39). Hybridization with polyadenylated RNA extracted from lactating mammary gland served as a control. ]mm unofluorescence: Mammary ceils recovered from collagen gels and labeled for I h with [~H]thymidine were cytocentrifuged on glass sides, fLxed briefly in ether-ethanol (1:1), reiiydrated with Phosphate-buffered saline (PBS), and reacted with goat anti-rabbit milk protein sera (7), diluted 1:10 (antibody titer ~0.05-0.1 mg/ml). The slides were washed three times in PBS, exposed for 15 min to biotinylated sheep F(ab')2 directed against goat IgG, and £mally incubated with lissamineor FITC-labeled streptavidin (R. Rodewald, D. Papermaster, and J. P. Kraehenbuhl, manuscript in preparation). The slides were then washed three times in PBS and once in distilled water, and were processed for autoradiography with fiquid Ilford L4 emulsion. The autoradiograms were exposed for 2 wk, developed, and then observed in a Zeiss photomicroscope II with a fluorescent attachment and equipped with Osram HBO 100-W high pressure mercury vapor light source, BGI2 excitation filter, and BG38 suppression filter.