The chick oviduct in tissue culture

Experimental Cell Research 15.5(1984) 252-260

The Chick Oviduct II. Estrogen

in Tissue Culture

Affects Ovalbumin Synthesis Oviduct Cell Proliferation

Differently

than

SALLY S. SEAVER,‘, * STEPHEN M. BAIRD2 and DEBRA F. SKAFAR’ ** ‘Department of Molecular Biology, Vanderbilt University, Nashville, TN 37235, ‘Veterans Administration Medical Center, San Diego, CA 92161 and Department of Pathology, The General Clinical Research Center, University of California, San Diego, CA 92093, USA

The addition of estradiol alone to oviduct cell cultures was sufftcient to induce ovalbumin synthesis, detectable both by immunofluorescence and immunoprecipitation of newly synthesized protein. Most cells stained positively for ovalbumin indicating that the culture conditions promoted the growth of the ovalbumin synthesizing tubular gland cells relative to other cell types. The rate of ovalbumin synthesis was lower than that expected in vivo but as high as or higher than that found in organ culture. In tissue culture ovalbumin gene expression was under the direct influence of estrogen. Previous work showed that estrogen did not stimulate rapid proliferation of oviduct cell cultures (S. S. Seaver, J. van der Bosch & G. Sato, Exp cell res 155 (1984) 241) [5]. Therefore further experiments were done in vivo to correlate the effects of different hormonal regimes on oviduct growth and ovalbumin synthesis. In several instances the hormones affected oviduct growth differently than they affected ovalbumin synthesis. However, there was a strong correlation between the ability of a hormonal regime to stimulate oviduct growth in vivo and the ability of the serum from those chicks to stimulate oviduct cellular proliferation in culture. In vivo estrogen also stimulates oviduct growth by very different mechanisms than it stimulates the expression of the egg white protein genes. 0 1964 Academic Press, Inc.

There are three types of epithelial cells in the oviduct of mature laying hens. Ciliated cells, which help propel the egg yolk down the lumen, and goblet cells, which synthesize avidin, line the lumen [l]. The much larger tubular gland cells, which comprise -70% of the mature oviduct and are connected to the lumen via ducts, are the site of synthesis of most egg white proteins such as ovalbumin, conalbumin, ovomucoid and lysozyme [2, 31. In the immature chick estrogen treatments alone stimulate both the growth and differentiation of the oviduct and the synthesis of egg white proteins, especially ovalbumin [l, 41. We have shown that chick oviduct cells can be induced to proliferate in culture and that rapid proliferation is under endocrine control [5]. We were interested in identifying the types of cells that were proliferating in culture and determining if * To whom offprint requests should be sent. Please request reprint numober RP086. Current address: Millipore Corporation, 80 Ashby Road, Bedford, MA 01730, USA. ** Current address: Department of Radiation Biology and Biophysics, University of Rochester Medical Center, Rochester, NY 14642, USA. Copyright @ 1984 by Academic Press, Inc. All rights of reproduction in any form reserved COl4-4827/84 $03.00

Estrogen

effects on growth us gene expression

253

egg white protein synthesis was under endocrine control. Since -50% of all protein in the laying hen oviduct is ovalbumin [ll, and it can be induced in immature chicks by estrogen treatments [2], ovalbumin is a good marker for tubular gland cells and for studying endocrine control. In this paper, we report on the ability of estrogen to induce ovalbumin synthesis in oviduct tissue cultures. Using both in vivo and tissue culture experiments we have compared the effects of estrogen on the synthesis of ovalbumin with its effects on oviduct growth. We have found that estrogen regulates oviduct growth by different mechanisms than it regulates egg white gene expression.

MATERIALS

AND METHODS

We initiated the primary stimulation of female chicks of the White Leghorn, DeKalb, Rhode Island Red or Black Rock Strains by implanting diethylstilbestrol (DES) pellets (Vineland Laboratories, Vineland, N.J.) in the neck or by injecting daily 0.5 mg of DES dissolved in sesame oil into the leg muscle. Chicks which had had DES pellets removed at least 2 weeks previously were considered to be withdrawn chicks. Secondary stimulated chicks are withdrawn chicks which have been restimulated with an estrogen. Immunofluorescent localization of ovalbumin was done by a modification of the procedure described by Baird [6]. Cultures were plated in 5 % horse serum in MCDB 201 [7] in the center of 60-mm collagen-treated plastic dishes by placing sterile 20 mm diameter stainless steel rings in them [5]. After 2 days the rings were removed, and cells were grown for an additional 4 days on the desired media. Cells were washed, fixed in methanol or 70% ethanol (preferred fixative) and stored at 4°C until use. Attempts to grow cells on collagen-treated glass coverslips met with only limited success. The area to be stained was wetted with 0.5 % bovine serum albumin (BSA) in phosphate-buffered saline (PBS: 8 g/l NaCl, 0.2 g/l KCI, 1.15 g/l Na2HP04, 0.2 g/l KH,PO,). Sixty microliters of the appropriate dilution of sheep anti-ovalbumin [8] was incubated with the fixed cells for 30 min at 37°C. Dishes were rinsed and incubated with two changes of PBS, 15 min each. FITC rabbit anti-sheep gamma globulin (MilesYeda) diluted 1 : 8 in 0.5 % BSA in PBS was incubated with the dishes for 60 min at 37°C. Dishes were rinsed with PBS and incubated with PBS for two changes of 30 min each. They were drained, counterstained with five drops of 0.06% Evans Blue in PBS for 10 set, rinsed three times in PBS, drained and five drops of 10% PBS (pH 8.5) in 90% glycerol was added. Plates were stored in the refrigerator in the dark. The use of the pH 8.5 PBS on the plates increased the intensity of the fluorescein fluorescence. Non-fluorescent cells could be identified since the Evans Blue made them red when the dye was excited with the mercury lamp. Cells were photographed using Kodak Ektachrome ASA 400 EL 135 on a Zeiss microphoto 2 microscope. Black and white photographs were made from the color slides using a No. 58 green filter, which reduced the passage of red light. Newly synthesized ovalbumin was identified by labeling new synthesized proteins with [3H]leucine and immunoprecipitating the ovalbumin. Cultures were plated and grown in MCDB 201 containing different sera at 5 % and different levels of estradiol (Calbiochem Behring). Dishes were washed twice with regular medium, once with medium lacking leucine, and incubated in 1.0 ml of MCDB 201 lacking leucine. After 10 min, 1.5uCi of [3H]leucine (ICN, 40-55 Ci/mmole) was added and incubation was continued for 1 h at 40°C. Cells were washed several times in PBS and then were lysed in a buffer containing NP-40 following the procedure of van der Bosch et al. [9]. The amount of labeled ovalbumin was determined by direct immunoprecipitation with sheep anti-ovalbumin according to the procedure of Seaver [lo]. In vivo rates of ovalbumin synthesis were determined by labeling oviduct minces and immunoprecipitation as previously described [lo]. The ovalbumin used to immunize the sheep was purified by DEAE-cellulose chromatography [ 111. The antibody containing fraction of the sheep serum was enriched by a 35% ammonium sulfate precipitation. Ouchterlony tests showed no cross reaction with conalbumin, ovomucoid or lysozyme. The protein immunoprecipitated by the sheep anti-ovalbumin was a single band which migrated with ovalbumin on SDS-PAGE (results not shown). Exp Cell Res IS5 (1984)

254

Seaver, Baird and Skafar

RESULTS Ovalbumin

Gene Expression

in Tissue Culture

Ovalbumin gene expression in culture was assessed by two different techniques, by immunofluorescence and by immunoprecipitation of newly synthesized protein. Immunofluorescence helps to determine the fraction of the cells synthesizing ovalbumin. Immunoprecipitation quantifies ovalbumin synthesis. Cells that were cultured in 5% horse serum containing 100 nM estradiol and stained with the sheep anti-ovalbumin and fluorescein-conjugated rabbit antisheep IgG had cytoplasms which stained the green-yellow color characteristic of fluorescein (fig. 1A). The staining was most intense around the nucleus. This pattern has been interpreted as being due to antigen bound to rough endoplasmic reticulum [12]. Such staining would be expected for ovalbumin which is synthesized on membrane-bound polysomes [13, 141.The nuclei did not fluoresce. They were light red due to the Evans Blue counterstain. No fluorescence was observed in cells cultured on serum alone (no estradiol present) or in cells cultured with estradiol but stained only with the fluorescein-conjugated rabbit anti-sheep IgG (fig. lD, E). Fluorescence due to the presence of ovalbumin was detected when cells were incubated with concentrations of estradiol less than 100 nM. The exact concentration of estradiol needed to visualize fluorescence due to intracellular ovalbumin varied from experiment to experiment. In general positive results were obtained when estradiol concentrations were greater than 50 nM. This is the same concentration of estradiol needed to induce maximum synthesis of ovalbumin in organ culture [ 171.Higher concentrations of estradiol(l50 nM) could also induce ovalbumin synthesis. Even cell cultures started from withdrawn chick oviducts could be induced to synthesize ovalbumin when estradiol was added (fig. 1B). No ovalbumin was detected in these cells before the estradiol was added (results not shown). Most of the cells contained ovalbumin even though at the time of seeding the proportion of ciliated cells was higher in withdrawn oviduct cultures than in stimulated oviduct cultures and there was little proliferation in horse serum-containing medium [5]. This may be further evidence that the epithelial cells of the oviduct can interconvert [ 151. Rarely were there cells that did not fluoresce after the estradiol was added. Only in one experiment were there patches of cells that did not fluoresce (fig. 1 C, bottom center extending up and towards the left). In this experiment cells were plated at a much higher density than normal so there was little room for growth. This was also the only time we observed ovalbumin sequestered in secretory granules. These cultures were started from chicks after 7 days of secondary stimulation. Such large secretory granules have been observed in chick oviducts after prolonged hormonal stimulation [16] and in laying hen oviducts just before the yolk enters the oviduct magnum 1171. Exp Cell Rcs 155 (1984)

Estrogen effects on growth vs gene expression

255

I. Immunofluorescent localization of ovalbumin in cell cultures. Cells were fixed, incubated with sheep anti-ovalbumin and fluorescein-conjugated rabbit anti-sheep IgG and counterstained with Evans Blue. The Evans Blue stains non-reactive cells and nuclei red. A No. 58 green filter converted the green-yellow fluorescence to light grey-white and the red Evans Blue stain to faint grey. (A) Cells from oviducts of chicks implanted for 7 days with DES pellets grown in 5 % horse serum plus 100 nM estradiol for 4 days. (B) Cells from withdrawn chick oviducts grown in 5 % horse serum plus 100 nM estradiol. (C) Cells from oviducts of chicks after 7 days of secondary stimulation grown in 5% charcoal-treated horse serum plus 100 nM estradiol. There are cells that did not stain yellow-green in the dark area-lower center edge extending up to the left. (0) Control culture treated the same as sample A except no estradiol in growth medium. (E) Control culture treated the same way as sample C except not incubated with the sheep anti-ovalbumin. Sample E was incubated with the fluoresceinconjugated rabbit anti-sheep IgG antibody. The magnification is the same for all pictures. White bar in @I, 5 pm.

Fig.

Ovalbumin synthesis in both estrogen-stimulated and withdrawn cultures was also detected by pulse-labeling the cells with [3H]leucine and immunoprecipitating the newly synthesized ovalbumin (table 1). These experiments corroborated the results obtained by immunofluorescence. Growing cells in serum alone resulted in very little, if any, ovalbumin synthesis. The addition of estradiol increased ovalbumin synthesis. The amount of ovalbumin synthesis in culture was lower than that measured in vivo after 4-6 days of stimulation. Similar reduced synthesis of ovalbumin was observed in organ cultures [18], where it was reported that somatomedin-like peptides act synergistically with estradiol to increase ovalbu17-848341

Exp Cell Res IS.5 (1984)

256 Seaver, Baird and Skafar Table 1. Ovalbumin synthesis in oviduct cell culture Source of oviducts

Days in culture

Time on estradiol

Estradiol cont. (nM)

Serum used”

Ovalbumin synthesisb (p/o)

14 days primary stimulated chick (DES pellets)

8 8 5 5

50 h 50 h 4 days 4 days

25 25 100 100

HS WCS cHS cscs

1.7kO.5 1.5kO.3 4.4kO.9 7.9kO.3

4 days 4 days 4 days 4 days

0 30 3o 30

HS HS WCS scs

0.6kO.4 l.OZbO.2 3.7kO.7 3.8kO.2

Withdrawn chick

a Concentration of estrogen-stimulated b % of total protein immunoprecipitating

serum used was 5 %. Sera used were from horse (HS), withdrawn chicks (WCS), chicks (SCS). cHS and cSCS are charcoal-treated sera. synthesized as determined by labeling cells with [3H]leucine in culture for 1 h and with sheep antiovalbumin [14]. All samples were assayed in triplicate.

min mRNA levels [19]. The presence of factors that act synergistically with estrogen may also explain why ovalbumin synthesis was usually higher in cultures maintained in chick serum than in cultures maintained in horse serum at the same concentration of estradiol. Effects of Estrogen on Growth us Gene Expression Further evidence that estrogen controls oviduct growth by different means than it controls gene expression was shown by a combination of in vivo and in vitro experiments. Newborn female chicks were injected with 0.5 mg of DES per day. After 14 days chicks were injected for 10 more days with 5 mg/day of DES, 5 mg/day of estradiol, continued on 0.5 mg/day of DES, switched to 5 mg/day of estradiol or administered no hormone at all. During the 10 days oviduct weights and ovalbumin synthesis were determined. At the end of the 10 days with the second hormone, serum from each group of chicks was collected. The ability of these sera to promote oviduct cellular proliferation in culture was ascertained. The effects of the different in vivo hormone treatments on oviduct weights in vivo was not surprising. The highest levels of estrogen caused the largest weight gain (fig. 2). Both estradiol benzoate and DES had similar effects. The oviducts grew less rapidly in chicks treated with only 0.5 mg/day of estradiol benzoate. Maintaining chicks on 0.5 mg/day of DES resulted in no significant increase in oviduct weight. As chicks get older and larger it takes more hormone to increase oviduct weight in vivo. Cessation of hormone treatment caused a 73 % decrease in oviduct weight by the end of the experiment. The effects of the serum collected from each group at day 10 on oviduct cellular proliferation in culture was very similar (fig. 3). Cells grown in sera from chicks stimulated with 0.5 mg/day or 5.0 mg/day of estradiol benzoate or 5.0 mg/day of Exp Cd

Res 155 (1984)

257

Estrogen effects on growth us gene expression

I 0

2 DAYS

4

6

OF HORMONE

8

IO

TREATMENT

0

4

2 DAYS

ON

EACH

SPECiFiC

6 SERUM

2. Effects of different estrogen treatments on oviduct wet gain in vivo. All chicks were administered 0.5 mg DES/day for 14 days. At day 0 chicks were divided into five groups which then received O-O, 5.0 mg/day of DES; 0-Q 5.0 mg/day of estradiol benzoate; W-D, 0.5 mg/day of estradiol benzoate; O-0, 0.5 mg/day of DES; or A-A, no hormone. Oviduct weights shown represent the average of at least three values. Fig. 3. Effects of sera from chicks receiving different estrogen treatments on oviduct cell proliferation in culture. Cell cultures were started from oviducts of chicks which had been implanted with DES pellets for 7 days. Cells were allowed to attach for 2 days in media containing 5 % horse serum before changing to the different sera. Oviduct cell culture grown in serum (5%) from chicks which had received 14 daily injections of 0.5 mg DES followed by 10 daily injections of 0, 5.0 mg estradiol benzoate; 0, 5.0 mg of DES; n , 0.5 mg of estradiol benzoate; 0, 0.5 mg of DES or A, vehicle only. Media was changed every two days and cell number was determined in duplicate by counting nuclei. All errors not shown were less than the size of the symbol. Fig.

DES showed the greatest proliferation, 13-14-fold. Cells grown in serum from chicks stimulated with low levels of DES showed less proliferation, a 6-7-fold increase. Cells grown on serum from vehicle-treated chicks showed little growth. The rate of oviduct growth in vivo was a fairly good predictor of the ability of serum from that chick to promote cellular proliferation in culture. However, the effects of the different estrogen treatments on oviduct growth were not always useful in predicting their effects on ovalbumin synthesis in vivo. At day 10 ovalbumin synthesis in the chicks maintained on 0.5 mg/day of DES was similar to that measured in chicks at day 0 (fig. 4). In chicks which had received no hormone for 10 days, there was no ovalbumin synthesis. Both of these results were expected. Quite unexpected were the insignificant amounts of ovalbumin synthesis in chicks which had received 5 mg/day of either DES or estradiol benzoate or the barely significant amounts of ovalbumin synthesis in chicks which had received 0.5 mg/day of estradiol benzoate. We have no explanation for why the higher doses of steroid inhibited egg white protein expression. Exp Cell Res 155 (1984)

258 Seaver, Baird and Skafar

Fig. 4. Effects of the various estrogen treatments on ovalbumin synthesis in vivo. Chicks were

stimulated in vivo with daily injections of 0.5 mg DES for 14 days. At day 0 of the experiment chicks were divided into five groups which received 0.5 mg or 5.0 mg of estradiol benzoate, 0.5 or 5.0 mg of DES or vehicle as outlined in fig. 2. At day 10 ovalbumin synthesis in vivo was determined by labeling organ minces with [3H]leucine and immunoprecipitating the newly synthesized ovalbumin in triplicate [lo]. The rate of ovalbumin synthesis is expressed relative to total cellular protein synthesis. The hatched bar shows the rate of ovalbumin synthesis at day 0.

DISCUSSION The addition of estradiol alone was sufficient to induce ovalbumin synthesis in cultures started from oviducts of primary stimulated, withdrawn or secondary stimulated chicks. As shown by immunofluorescent labeling, ovalbumin was present in most cells. The culture conditions seem to be promoting the growth of tubular gland cells, the cells responsible for making ovalbumin. Ovalbumin synthesis could also be detected by labeling the culture with [3H]leucine and immunoprecipitating the newly synthesized ovalbumin. The amount of ovalbumin synthesis after 24 days, l-8 % of total protein synthesis, is lower than what would be found in vivo by 2-4 days of stimulation [2, 201. The amount of estradiol needed to induce ovalbumin synthesis, 25-100 nM, is much greater than the amount needed to saturate the receptor (l&10 nM). The need for 10-100 nM estradiol to induce similar levels of ovalbumin synthesis was also reported in ,,viduct organ cultures [18]. There are at least two reasons for these differences. First additional factors besides estrogen are probably required for ovalbumin synthesis. The addition of a somatomedin-like peptide to oviduct organ cultures containing estradiol increases ovalbumin mRNA lo-fold [19]. In the quail low doses of estrogen do not induce ovalbumin synthesis [21, 221. Higher doses do because, it is hypothesized, the estrogen has a limited ability to function as progesterone which also leads to the sequestering of ovalbumin in secretory granules [21]. In the chick either progesterone or testosterone is needed in addition to estrogen to sequester ovalbumin in secretory granules [23]. Second, when oviduct cells are taken from an estrogen stimulated chick for culture, they Exp Cell Res 155 (1984)

Estrogen effects on growth us gene expression

259

are forced into withdrawal since for the first 2 days in culture there is very little exogenous estradiol [5]. Changing the levels of estradiol administered in vivo can have very marked and often inhibitory effects on ovalbumin synthesis (fig. 4, [24]). Similar phenomena may be occurring in cell or organ culture. Both in vivo and in culture oviduct cellular proliferation and ovalbumin synthesis are under endocrine control. However, the mechanism of endocrine control is very different for the two processes. Estradiol is sufficient to stimulate ovalbumin synthesis in culture. Rapid cellular proliferation in culture depends on factors present in the serum from chicks with rapidly growing oviducts [5]. The ability of a serum to stimulate cellular proliferation in culture correlates well with the rate of oviduct growth. However, even in vivo it is possible to uncouple the stimulatory effect of estrogen on oviduct growth (fig. 2) from its effect on ovalbumin synthesis (fig. 4). In most cases different factors are involved in promoting growth of an organ than are involved in promoting the differentiated functions of that organ. The chick oviduct is one of the few organs in which the same factor, estrogen, has been implicated in both growth and gene expression. Our work using a combination of tissue culture and in vivo experiments strongly suggests that the detailed mechanism by which estrogen stimulates oviduct growth are very different from the ways it stimulates egg white protein gene expression. This research was supported by the Veterans Administration (S. M. B.) and NSF (grant No. PCM 79-23025 to S. S. S.), Vanderbilt University Research Council (S. S. S.), Vanderbilt University Population Center (NICHD-07043 to D. F. S.). P. B. Coulson and G. Mosig are thanked for their helpful comments and Susan Reavis and Syble Mitchell are thanked for their excellent and quick secretarial assistance.

REFERENCES 1. O’Malley, B W, McGuire, W L, Kohler, P 0 & Korenman, S G, Recent prog horm res 25 (1969) 10.5. 2. Palmiter, R D, J biol them 247 (1972) 6450. 3. Palmiter, R D & Gutman, G A, J biol them 247 (1972) 6459. 4. Schimke, R T, Pennequin, P, Robins, D & McKnight, G S, First European symposium of hormone and cell regulation (ed J Dumont & J Nunez) p. 209. Elsevier, Amsterdam and New York (1977). 5. Seaver, S S, van der Bosch, J & Sato, G, Exp cell res 155 (1984) 241. 6. Baird, S, Mol immunol 17 (1980) 959. 7. McKeehan, W L & Ham, R G, J cell bio171 (1976) 727. 8. Tokarz, R R, Harrison, R W & Seaver, S S, J biol them 254 (1979) 9178. 9. van der Bosch, J, Masui, H & Sato, G, Cancer res 41 (1981) 611. 10. Seaver, S S, J steroid biochem 14 (1981) 949. 11. Mandeles, S, J chromatogr 3 (1960) 256. 12. Bergmann, J E, Tokuyasu, K T & Singer, S J, Proc natl acad sci US 78 (1981) 1746. 13. Rhoads, R E, McKnight, G S & Schimke, R T, J biol them 248 (1973) 2031. 14. Palmiter, R D, Christensen, A K & Schimke, R T, J biol them 245 (1970) 833. 15. Richardson, K C, Royal sot London phi1 trans ser B 225 (1935) 149. 16. Kohler, P 0, Grimley, P M & O’Malley, B W, J cell biol 40 (1969) 8. 17. Wybum, G M, Johnston, H S, Draper, M H & Davidson, M F, J exp physiol 55 (1970) 213. 18. McKnight, G S, Cell 14 (1978) 403. Exp Cell Res 155 (1984)

260 Seaver, Baird and Skafar 19. 20. 21. 22. 23.

Evans, M I, Hager, L J & McKnight, G S, Cell 25 (1981) 187. Palmiter, R D & Wrenn, J T, J cell biol50 (1971) 598. Sandez, D, Boisvieux-Ulrich, E, Laugier, C & Brard, E, Gen camp endocrinol26 (1975) 451. Laugier, C, Pageaux, J-F, Soto, A M & Sonnenschein, C, Proc natl acad sci US 80 (1983) 1621. Nalbandov, A V, Reproductive physiology of mammals and birds, 3rd edn, p. 153. W H Freeman & Co., San Francisco (1976). 24. Skafar, D F & Seaver, S S, Endocrinology. In press. Received March 1, 1984 Revised version received May 3, 1984

Exp Cell Res IS5 (19.34)

Printed

in Sweden