JOURNAL OF CELLULAR PHYSIOLOGY 145:333-339 (1990)

Estrogen Regulates Peptidylarginine Deiminase Levels in a Rat Pituitary Cell Line in Culture SABURO NACATA,* MlHO YAMAGIWA, KINJIINOUE, AND TATSUO SENSHU

Department of Biochemistry, Tokyo Metropolitan Institute of Gerontology, Tokyo 173, japan (S.N., M.Y., T.S.); Department of Morphology, Institute of Endocrinology, Cunma University, Maebashi, lapan fK.1) A Nonidet P-40 extract of growth hormone-producing rat pituitary MtTlS cells was found to contain peptidylarginine deiminase (EC 3.5.3.15), which was indistinguishable from an enzyme preparation from rat muscle in Western immunoblotting and immunoprecipitation. This enzyme was immunocytochemically detected in the cytoplasm but was not secreted into the medium during the cultivation. When the cells were cultured for 2 days with various concentrations of 17P-estradiol (E,), the enzyme activity increased in a dose-dependent manner, reaching a maximum level (four- to fivefold higher than control)at about 1 Op9 M. This increase in the enzyme activity was evident by 14 hr of culture and became relatively stable after 24 hr. It correlated well with the increase in the amount of the muscle type enzyme per cell as analyzed by Western immunoblotting. Estriol

and a synthetic estrogen, diethylstilbestrol, also increased the enzyme activity, whereas testosterone, progesterone, and corticosterone were without effect. An antiestrogen, tamoxifen, which by itself was inactive, partially suppressed the effectof E,. Exposure of MtT/S cells for 14 hr to E, increased incorporation of 35S-labeledamino acids into the immunoprecipitable peptidylarginine deiminase. This increase was dependent on the concentration of E,, attaining a maximum level (about tenfold higher than the control) at about lo-’ M. These results indicate that estrogen effects the increase in peptidylarginine deirninase content in the pituitary cells by stimulating enzyme synthesis.

Recent studies have shown that wide varieties of mammalian tissues have peptidylarginine deiminase (protein-L-arginine deiminase; EC 3.5.3.15) that catalyzes the conversion of L-arginine residues in proteins to L-citrulline residues (Rogers et al., 1977; Kubilus et al., 1980; Fujisaki and and Sugawara, 1981; Kubilus and Baden, 1983; Takahara et al., 1983,1989; Watanabe et al., 1988; Nagata and Senshu, 1990). The enzyme proteins have thus far been purified to apparent homogeneity from bovine brains (Kubilus and Baden, 1983) and skeletal muscles of rabbit (Takahara et al., 19831, rat (Watanabe et al., 19881, and mouse (Takahara et al., 1989).In addition to this muscle-type enzyme, the presence of at least two other distinct types of eptidylarginine deiminase, designated epidermal an hair follicle types, has been suggested from differences in their substrate specificities, immunochemical properties, and tissue distributions (Watanabe et al., 1988). The enzyme action in vitro has been shown to cause changes in physicochemical and biological properties of certain protein substrates (Takahara et al., 1985; Inagaki et al., 19891, suggesting important physiological roles of the enzyme in vivo. In fact, the enzymes present in mammalian hair follicle (Rogers et al., 1977) and epidermis (Kubilus et al., 1980; Fujisaki and Sugawara, 1981) appear to be involved in the deimination of arginine residues in precursor proteins to form unique structural proteins present in these tissues (Rothnagel and Rogers, 1984).In addition,

B

0 1990 WILEY-LISS, INC

citrulline-containing forms of myelin basic proteins have been shown to exhibit altered lipid-aggregating properties (Wood and Woscarello, 1989). However, physiological roles of the enzyme in many other tissues remain unknown, mainly because citrulline-containing reaction products have not been identified in these tissues. In the revious paper, we report marked sex difference in t e skeletal muscle-type peptidylarginine deiminase content in rat pituitaries (Senshu et al., 1989). It increases in female pituitaries with sexual maturation and shows cyclic changes around the estrous cycle. The results of ovariectomy and estrogen administration experiments suggested that the pituitary enzyme content is under direct or indirect control of estrogen. Our group has also shown that the enzyme immunoreactivity was localized in specific hormone-producing cells in female rat pituitaries (Akiyama et al., 1989). We are interested in the mechanisms regulating peptidylarginine deiminase levels in relation to its possible involvement in the functional activity of the hormone-producing cells.

R

Received March 19, 1990; accepted August 3, 1990. *To whom reprint requestsicorrespondence should be addressed. Miho Yamagiwa’s present address is Banyu Pharmaceutical Go., Nihonbashi Honcho-2-2-3,Chuoku, Tokyo-103, Japan.

334

NAGATA ET AL.

In the present study, we used an in vitro culture system to examine whether estrogen affects the peptidylarginine deiminase level of pituitary cells. Similar systems have been used successfully for investigating the mechanisms regulating cell growth, hormone production, and hormone secretion of pituitary cells at cellular and molecular levels (Lieberman et al., 1978; Vician et al., 1979). A newly established rat pituitary cell line, MtT/S (Inoue et al., 1990), used here, allowed the large-scale cultures required for determining enzyme activities. We present here evidence that exposure of these cells to physiological concentrations of estrogen results in increase of the eptidylarginine deiminase content through the stimu ation of enzyme synthesis.

P

MATERIALS AND METHODS Media and steroids Dulbecco’s modified Eagle’s medium (DMEM; containing 1 g/liter glucose) and Ham’s F-12 nutrient mixture (F-12) were purchased from GIBCOlBRL Life Technologies, Inc. (Gland Island, NY) and Flow Laboratories (Englewood, CAI, respectively. Eagle’s minimum essential medium (MEM) free of phenol red was from Nissui Pharmaceutical Co. (Tokyo, Japan). Methionine-free MEM was prepared in our laboratory according to the recipe of Nissui Pharmaceutical Co. Fetal bovine serum, normal horse serum, &enkillin, streptomycin and L-glutamine were from G CO/BRL Life Technologies, Inc. Other amino acids, vitamins, and other medium components were from Takara Kosan Co. (Tokyo, Japan), Sigma Chemical Co. (St. Louis, MO), and Wako Pure Chemical Industries (Osaka, Japan). Tissue culture-grade 17P-estradiol (EJ, testosterone, ro esterone, and rea ent-grade corticosterone, estrior and diethylsti bestrol (DES) were obtained from Sigma Chemical Co. They were dissolved in ethanol at 1 mM and stored at -7O’C. Tamoxifen (Si a Chemical Co.) was dissolved at 100 mM in ethano immediately before use. In some cases, sera were preabsorbed with dextran-coated charcoal [200 mg dextran T-75 (PharmacidLKB Biotechnology, Inc., Uppsala, Sweden) and 1g charcoal powder (Wako Pure Chemical Industries)/100 ml serum] to deplete steroids according to Lagace et al. (1980). Cell culture A clonal diploid cell line, MtTIS, was established from the mammotropic pituitary tumor, MtTIF84, that had been induced in a Fisher F344 rat by continuous estrogen stimulation (It0 et al., 1985; Inoue et al., 1990). The cells were maintained in a 1:l mixture of DMEM and F-12 supplemented with 10%normal horse serum, 2.5% fetal bovine serum, penicillin, stre tomycin sulfate, 3 g/liter glucose, and 2.85 glliter Na C03 in 5% COz and 95% air at 37°C. Under these culture conditions, MtT/S cells produce and secrete growth hormone (GH), but not prolactin (PRL), into the medium (Inoue et al., 1990).They were subcultured every 5-7 days after dislodging from the substratum by gentle pipetting. To examine the effects of steroids, cells were precultured for 7 days in henol red-free MEM supplemented with dextran-cYlarcoal-treated sera. After the preculture, the cells were harvested,

(If3),

f

!?

2

counted using a hemocytometer, and resuspended at appropriate concentrations in the fresh medium. Cultures containing diluted steroids were set up at l x lo7 cellsll5 ml in plastic dishes (100 X 20 mm; Corning Glass Works, Corning, NY). Control cultures were incubated with the same concentration (0.01% or less) of ethanol, a concentration we found to have no effect on cell growth or the enzyme content. Enzyme assay Cells were harvested by pipetting, counted, and washed once with phosphate-buffered saline (10 mM sodium phosphate, 150 mM NaC1, pH 7.2; PBS). In some experiments, pelletted cells were stored at - 70°C. The cells were resuspended at lo8 cellslml in the lysis buffer [50 mM Tris HC1, pH 7.6,lOO mM NaC1,lO mM 2-mercaptoethanol, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 0.25% (wt/vol) Nonidet P-401 and incubated for 30 min on ice with occasional vortexing. After the incubation, the tubes were centrifuged for 10 min at 10,OOOgand the supernatants were saved for the enzyme assay. Peptidylarginine deiminase activity in the cell extracts were determined using N-benzoylL-arginine-0-ethylester (Nakalai Tesque Inc., Kyoto, Japan) as a substrate with the method described elsewhere (Senshu et al., 1989). The enzyme assay after immunoprecipitation was performed as described previously (Nagata and Senshu, 1990). One unit of the enzyme was defined as the amount of enzyme required to catalyze the deimination of 1pmole substrate in 1hr at 50°C. Immunofluorescence staining of cells Cells were cultured on poly-L-lysine-coated Lab-Tek tissue culture slides (Nunc, Inc., Naperville, IL). After brief washing in PBS, the cells were fixed in ice-cold acetone:methanol (1:l)for 20 min. They were treated with 1%normal goat serum before probing with a rabbit antirat skeletal muscle peptidylarginine deiminase antiserum (1:2,000 dilution in PBS containing 1% bovine serum albumin) for 2 hr at room temperature. This antiserum has shown to react specifically with the muscle-type enzyme (Watanabe et al., 1988). The cells were then washed overnight in PBS containing 0.1% Tween 20 at 4°C and reincubated for 2 hr at room temperature with a fluorecein-conjugated antirabbit IgG goat antibody (Tago, Inc., Burlingame, CA; 1:lOO dilution in PBS containing 1%normal goat serum) for 2 hr. After washing overnight, the cells were mounted with PBS-glycerol (3~1) and observed with a fluorescence microscope with epiillumination. Western immunoblot analysis Western immunoblot analysis of cell extracts was performed as described previously (Senshu et al., 1989). The antirat skeletal muscle peptidylarginine deiminase rabbit antiserum was used at a dilution of 1:2,000. Rabbit antisera against rat GH (HAC-RC-25-01RBP85) and rat PRL (HAC-RT-26-01-RBP85)were kindly rovided by Dr. Wakabayashi of the Hormone Assay enter, (Institute of Endocrinology, Gunma University) and used at dilutions of 1:5,000 and 1:10,000, respectively. The horseradish peroxidase-conjugated antirabbit IgG goat antibody (Bio-Rad Laboratories, Richmond, CA) was used at 1:3,000.

8

ESTROGEN AND ENZYME SYNTHESIS IN PITUITARY CELLS

Biosynthetic labeling of proteins MtT/S cells were precultured for 7 days in phenol red-free MEM supplemented with dextran-charcoaltreated sera. The cells were then resuspended at 1 x lo6 celldm1 in the fresh medium containing E2 or ethanol alone (control) and plated at 3 muwell in plastic six-well plates (Corning Glass Works). After cultivation for 14 hr, the medium was aspirated, and each well was rinsed once with methionine-free MEM and then fed with 1 ml methionine-free MEM containing E2 (or ethanol) and 60 KCi Tran35S-label (ICN Radiochemicals, Irvine, CA). After an additional 3 hr culture, cells were harvested, washed four times with cold PBS, counted, and lysed in the lysis buffer at 1 x lo7 cells/ ml. Incorporation of labeled amino acids into total cellular proteins was measured as described previously (Mishell and Shiigi, 1981). Briefly, 1 ~1 of the lysate was mixed with 500 pl each of 0.5%fetal bovine serum and 10% trichloroacetic acid containing 5 mM L-methionine and incubated for 30 min on ice. The precipitate was collected on a glass filter and washed with 5% trichloroacetic acid containing L-methionine, and the filter was dried and the radioactivity measured using a liquid scintillation counter. Immunoprecipitation The extracts of radiolabeled cells were normalized for the radioactivity of acid-precipitable proteins (Mishell and Shiigi, 1980)and precleared by incubations for 1hr on ice with a normal rabbit serum and then for 1 hr with protein A-Sepharose CL4B (PharmacialLKB Biotechnology, Inc.). The precleared extracts were immunoprecipitated with the antiserum against peptidylarginine deiminase, GH, and PRL for f3-16 hr at 4"C, followed by protein A-Sepharose CL-4B. For sodium dodecyl sulfate (SDSI-polyacrylamidegel electrophoresis, labeled proteins precipitated were extracted by boiling the beads for 5 min in Laemmli's sample buffer containing 2.3% SDS and 5 % 2-mercaptoethanol and were run on 10% polyacrylamide gels using the discontinuous buffer system (Laemmli, ,1970).Radioactive protein bands on the gels were visualized by fluorography using Enlightning (New England Nuclear, Boston, MA). To estimate rates of peptidylarginine deiminase biosynthesis, the labeled enzyme samples immunoprecipitated were mixed with the purified rat skeletal muscle enzyme (5 ng/p1) and electrohoresed. After staining with Coomasie brilliant blue, gands corresponding to the enzyme were cut out and incubated overnight at 37°C in a Protosol-Econofluor (New England Nuclear) mixture (1:20), and the radioactivities were determined with a liquid scintillation counter. Statistical analyses Significance of difference between the means of control and experimental groups was examined using the single-tailed Student's t test. RESULTS We have found significant levels of peptidylarginine deiminase activity in the extracts of MtTlS cells that were grown in the fully supplemented DMEM/F-12 mixture. The activity could be totally immunoprecipi-

335

tated with the antiserum to the rat skeletal muscle enzyme and protein A-Sepharose CL4B. Immunofluorescence staining using the same antiserum showed that the enzyme immunoreactivity was localized in the cytoplasm of most of the cells (Fig. 1). In addition, Western immunoblot analysis of the extract demonstrated a single protein band corresponding to the position of the purified rat skeletal muscle enzyme (Fig. 2). These results indicate that the peptidylarginine deiminase in MtT/S cells is the muscle-type enzyme. Exposure of cells to 10-12-10-7M E, in the fully supplemented DMEMIF-12 mixture did not alter their peptidylarginine deiminase activity (data not shown), so we precultured cells for 7 days in the phenol red-free MEM su plemented with dextran-charcoal-treated sera and t en exposed to E2 in the same medium. The preculture caused 4 0 4 0 % decrease in the enzyme activity. However, after this treatment, the cells responded to E2 with a si ificant increase in the enzyme activity. Figure 3A s ows enzyme activities in the extracts of cells that were cultured for 2 days with different concentrations of E,. The response was detectable at lo-" M and became greater with increasing concentrations of the steroid, attaining a maximum activity (four- to fivefold higher than the control) at lo-' M. Western immunoblot analysis of the corresponding extracts revealed a concurrent increase in the amount of the enzyme per cell (Fig. 3B). The enzyme activity in the extracts of E,-treated cells was totally immunoprecipitated with the antiserum and protein A-Sepharose CL4B. These results indicate that the increase in the enzyme activity is based on the increase in the amount of the skeletal muscle type enzyme per cell. In contrast, the steroid treatment did not alter the GH content. PRL was not detectable in cell extracts under any culture conditions examined (data not shown). No enzyme activity was detectable in the medium of either the steroid-treated or the control cultures. To test the specificity of the action of E,, effects of various steroids were compared under the same culture conditions. At a concentration of lo-' M, E, and diethylstilbestrol (DES) caused about a fourfold in-

i

!i?

Fig. 1. Immunofluorescence staining of MtTiS cells. The cells were cultured for 2 days in the fully supplemented DMEWF-12 mixture and stained with antirat skeletal muscle peptidylarginine deiminase antiserum and fluorecein-conjugatedantirabbit IgG antibody. Phasecontrast (A) and the fluorescence (B) micrographs of the same field. Bar = 10 km.

336

NAGATA ET AL.

crease in peptidylarginine deiminase activity (Fig. 4). E, was less effective, causing a smaller but significant increase at the same concentration. However, addition of M E, resulted in an increase of the enzyme activity comparable to that caused by lo-' M E, or DES. In contrast, testosterone, progesterone, and corticosterone had no significant effect at the concentrations tested. Next, we examined effects of an antiestrogen, tamoxifen, on the E,-dependent increase of the peptidylarginine deiminase contents according to a previously described procedure (Westley and Rochefort, 1980). The results showed that tamoxifen by itself did not affect the enzyme content at any concentration between lop9 M and M, a range of concentrations that showed no toxic effect on the cells (Fig. 5). However, it antagonized the effects of 10-l' M E2 in a dose-dependent manner, showing about 70%depletion of the increase of the enzyme activity at lop6 M. The time course of the increase in peptidylarginine deiminase activity in the cells was examined after the addition of lo-' M E2 into the cultures. The enzyme activity increased rapidly for initial 24 hr and then continued to show a further small increase up to 72 hr (Fig. 6A). This seemed to occur in parallel with the increase in the enzyme amount as demonstrated by Western immunoblot analysis (Fig. 6B).

04/ 0

,;-12

&lo 1;-9 1i-t Concentration of €2 (MI 1-;,

Fig. 3. Effects of different concentrations of E, on the peptidylarginine deiminase (PAD) content of MtTiS cells. Cells were precultured for 7 days in the phenol red-free MEM supplemented with dextrancharcoal-treated sera. The cells were then cultured for 48 hr with indicated concentrations of E,, and the extracts were prepared for the enzyme assay and Western immunoblot analysis. A Result of enzyme assay. Each point represents the mean value with SD (vertical bar) of triplicate cultures. An asterisk indicates that the value is statistically significantly different (P < 0.01)from the control value. B Results of Western immunoblot analysis. Lanes 1-6: Pooled samples at E, concentrations of O-lO-' M as in A.

OM

Fig. 2. Western immunoblot analysis of an extract of MtTiS cells grown in the fully supplemented DMEMIF-12 mixture. The extract from 1 x lo6 cells (lane 2) was electrophoresed along with 5 ng purified rat skeletal muscle peptidylarginine deiminase (lane 1) on a 10% polyacrylamide gel containing SDS. After the electrophoresis, proteins were transferred to a nitrocellulose membrane, and the enzyme on the membrane was visualized with a rabbit antiserum against rat skeletal muscle enzyme and a horseradish peroxidaselabeled antirabbit IgG antibody. The positions of molecular weight markers (Pharmacia) are indicated at right.

-

€3 T Pq cs Fig. 4. Effects of various steroids on the peptidylarginine deiminase (PAD) activity in MtTiS cells. Cells precultured as described in the legend to Figure 3 were cultured for 48 hr with the indicated concentrations of E,, E , DES, testosterone (T),progesterone (PJ, corticosterone(CS), or etkanol alone (control;C). Each bar represents the mean value of triplicate cultures with SD (vertical bar). An asterisk indicates that the value is statistically significantly different (P < 0.01) from the control value.

C

E2

DES

337

ESTROGEN AND ENZYME SYNTHESIS IN PITUITARY CELLS

0

o

10-9 10-8 10-7 10-6 Concentration of tamoxifen (M)

Fig. 5. Effect of tamoxifen on the increase of peptidylarginine deiminase (PAD) content of MtT/S cells by E,. Cells precultured as described in the legend to Figure 3 were cultured for 48 hr in media containing various concentrations of tamoxifen, and then the media were changed to those containing the same concentrations of tamoxifen alone (open circles) or tamoxifen plus lo-'' M E, (closed circles). After incubation for an additional 48 hr, cells were harvested and the enzyme activities determined. An asterisk indicates that the value is statistically significantly different (P< 0.01) from the value of the control (culture containing M E, alone).

To determine whether the increase in peptidylarginine deiminase content in the cells during cultivation with E2 was caused by increased enzyme synthesis, biosynthetic protein labeling experiments were performed. In these experiments, rates of incorporation of labeled amino acids into the enzyme protein were measured at a period when the enzyme content was increasing at the maximum rate. That is, cells were cultured for 14 hr with different concentrations of E,, and then they were pulsed for 3 hr with 35S-labeled amino acids in the continuous presence of the steroid. The amount of labeled amino acids incorporated into total cellular proteins was found to increase slightly by exposure to E2 in a dose-dependent manner (Fig. 7), showing about 40% increase at lo-' M. Next, the extracts of the labeled cells were immunoprecipitated and analyzed with SDS-polyacrylamide gel electrophoresis. Fluorographs of the gels demonstrated that the radioactivity incorporated into the immunoprecipitable enzyme appeared to increase with increasing concentrations of E2 added to the cultures (Fig. 8). Similar labeling experiments were conducted independently and the radioactivity incorporated into the enzyme was quantitated as described in Materials and Methods. The results summarized in Figure 7 confirm that the increase of the incorporation was dependent on concentrations of EP,attaining a maximum level (about a tenfold increase compared with the control and 0.01% of the incorporation into total proteins) at lo-' M. In contrast, E2 neither changed the incorporation into GH nor induced initiation of PRL synthesis at detectable levels (data not shown).

DISCUSSION The results presented above show that ex osure of MtT/S cells to physiological concentrations o estrogen results in an increase in their peptidylarginine deiminase contents. The other nonestrogenic steroid hormones tested had no effect. In addition, the antiestrogen tamoxifen suppressed the effect of E, in a manner as reported in secretory glycoproteins of human breast cancer cells (Westley and Rochefort, 1980), indicating that the response was estrogen specific. The sensitivity to E2and the time course of the response are similar to those of the production of secretory glycoproteins (Westley and Rochefort, 1980) and progesterone receptors (Kassis et al., 1984). However, this response seems more sensitive to E, than ovalbumin (McKnight, 1978) and PRL (Lieberman et al., 1978; Amara et al., 1987), which require more than lo-' M for maximum stimulation. Amara et al. (1987) have suggested that the different sensitivities of responses to estrogen could be explained at the level of mRNA accumulation, instead of distinguishable estrogen receptor molecules. Incorporation of radiolabeled amino acids into the immunoprecipitable enzyme increased after stimulation of the cells with E,. Since the enzyme is not secreted into the culture medium, the increase can be regarded as the result of increased rate of enzyme biosynthesis. Together with the observation indicating that MtT/S cells express the estrogen receptor (unpublished observation), the regulation may well involve the binding of the estrogen with its receptor molecule and activation of the rece tor's transcriptional regulatory function (Rories an Spelsberg, 1989). The next

P

I

I

48

72

1

Hours in culture

Fig. 6. Change in peptidylarginine deiminase (PAD)content during cultivation of MtTlS cells with E,. Cells precultured as described in the legend to Figure 3 were cultured with (closed circles) or without (open circles) lo-' M E, for the indicated times, harvested, and stored at -70°C for simultaneous assay. A Results of enzyme assay. Each point represents the mean value of triplicate cultures with SD (vertical bar). B Result of Western immunoblot analysis. Lanes 1-7 Pooled sample of 0 , 6 , 9 , 14, 24, 48, and 72 hr of culture.

f

338

NAGATA ET AL.

Fig. 7. Rates of incorporation of radiolabeled amino acids into the immunoprecipitable peptidylarginine deiminase (closed circles) and total proteins (open circles) of MtTiS cells that were cultured with different concentrations of E,. Cells precultured as described in the legend to Figure 2 were cultured for 14 hr with the indicated concentrationsof E,, and the rate of [35S]aminoacid incorporationinto the immunoprecipitable peptidylarginine deirninase was quantitated as described in Materials and Methods. Each point represents the mean value with SD (vertical bar) of triplicate cultures. An asterisk indicates that the value is statistically significantly different (P < 0.01) from the control value. Actual cprn in control culture was 60.3 x lo6 cpmiculture (total proteins) and 678 cpmiculture (irnmunoprecipitable enzyme fraction), with SD less than 20% of the mean values.

Fig. 8. SDS-polyacrylamide gel electrophoresis of immunoprecipitates from radiolabeled MtTiS cells. Cells precultured as described in the legend to Figure 3 were cultured for 14 hr with various concentrations of E, and then pulse-labeled for 3 hr using Trar~~~S-label (ICN Radiochemicals).Cell extracts were immunoprecipitated with antirat skeletal muscle peptidylarginine deiminase antiserum and protein A-Sepharose CL-4B and electrophoresed, and radioactive protein bands were visualized by fluorography. Positions of [14Clmolecular weight markers (New England Nuclear) are indicated on the right. Lanes 1 4 . samples at E, concentrations of 0, 10-l' lo-'' W 9 and , M. Arrowhead indicates the position of peptihylargi; nine deiminase.

question to be answered is whether the action of estrogen is directed toward transcription of the enzyme gene as in the case of other estrogen-induced proteins (McKnight, 1978; Vician et al., 1979; Amara et al., 1987). In a previous paper, we reported a sex difference, estrous cycle-related changes, and apparent estrogen dependence of the peptidylarginine deiminase content in rat pituitaries (Senshu et al., 1989). Recently, Takahara et al. (1989) reported similar estrous cycle-related changes (highest at estrous phase) of the enzyme activity in the mouse uterus. These facts and the fact that the plasma estrogen level increases at the proestrous phase (Butcher et al., 1974) suggested that the enzyme amount would be controlled by direct or indirect action of estrogen on these tissues. The present results support the idea that this occurs, at least in part, through the direct action of estrogen on target cells. Peptid larginine deiminase immunoreactivity was detectab e in the cytoplasm of some lactotrophs in female pituitaries but not in male lactotrophs or other hormone-producing cells (Akiyama et al., 1989).Therefore, it seems that the estrogen-dependent increase in the pituitary enzyme content (Senshu et al., 1989) is attributable either to proliferation of the enzymepositive lactotrophs or to increased content of the enzyme in individual lactotrophs. The present results support the latter possibility. However, the former mechanism could also partially contribute to the increase, since estrogen has shown to have a mitogenic activity on lactotrophs (Amara et al., 1987). Alternatively, conversion of other types of cells to the enzymepositive lactotrophs may be a mechanism. Evidence has been reported that both mammotrophs and lactotrophs are derived from GH-producing precursor cells (Bodner et al., 1988; Ingraham et al., 1988; Borrelli et al., 1989; Dolle et al., 1990). It should be noted that MtT/S cells, which did not produce detectable amounts of PRL in the present study, do initiate PRL production under certain culture conditions in vitro (unpublished observations). This suggests that MtTlS cells may be capable of differentiating into lactotrophs. It is tempting to postulate in this respect that the presence and the estrogen regulation of peptidylarginine deiminase in MtT/S cells represent a certain differentiation step of cells in lactotroph lineage. Based on the knowledge gained thus far, we hypothesize that the enzyme is involved in certain femalespecific function of pituitary hormone-producing cells. To evaluate this hypothesis, determination of actual target proteins of this enzyme is required. Recently, Inagaki et al. (1989) reported that this enzyme can deiminate arginine residues in intermediate cytoskeleta1 filament proteins in vitro in a micromolar Ca2+dependent manner and that it resulted in the loss of their polymerization competence. Based on these observations, the authors suggested that the enzyme may be involved in the regulation of the assembly-disassembly of cytoskeletal intermediate filaments. However, such a possibility has not yet been substantiated in vivo. ACKNOWLEDGMENTS We thank Dr. Watanabe and Mr. Akiyama of our laboratory for preparation of the rabbit antiserum

P

ESTROGEN AND ENZYME SYNTHESIS IN PITUITARY CELLS

against rat skeletal muscle peptidylarginine deiminase and Dr. Wakabayashi (Hormone Assay Center, Institute of Endocrinology, Gunma University, Gunma, Japan) for providing the rabbit antisera against rat GH and PRL. This work was supported in part by Science Research Grant 01580204 from the Ministry of Education, Science and Culture of Japan.

339

Kubilus, J., Waitkus, R.W., and Baden, H.P. (1980) Partial purification and specificity of an arginine-converting enzyme from bovine epidermis. Biochim. Biophys. Acta, 615246-251. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T,. Nature, 227:680-685. Lagace, L., Massicotte, J., and Labrie, F. (1980) Acute stimulatory effects of progesterone on lutenizing hormone and follicle-stimulating hormone release in rat anterior pituitary cells in culture. Endocrinology, 206:68p693. Lieberman, M.E., Maurer, R.A., and Gorski, J. (1978)Estrogen control of prolactin synthesis in uitro. Proc. Natl. Acad. Sci. USA, 7 5 5 9 4 6 LITERATURE CITED 5949. Akiyama, K., Inoue, K., and Senshu, T. (1989) Immunocytochemical McKnight, G.S. (1978) The induction of ovalbumin and nonalbumin mRNA by estrogen and progesterone in chick oviduct explant study of peptidylarginine deiminase: Localization of its immunorecultures. Cell, 24:403413. activity in prolactin cells of female rat pituitary. Endocrinology, Mishell, B.B., and Shiigi, S.M. (1981) Selected Methods in Cellular 225:1211-1227. Immunology. W. H. Freeman and Co., San Francisco, pp. 406-407. Amara, J.F., Van Itallie, C., and Dannies, P.S. (1987) Re lation of prolactin production and cell growth by estradiol: DiRrence in Nagata, S., and Senshu, T. (1990) Peptidylarginine deiminase in rat and mouse hemopoietic cells. Experientia, 46:72-74. sensitivity to estradiol occurs at levels of messenger ribonucleic acid Rogers, G.E., Harding, H.W.J., and Llesellyn-Smith, T.J. (1977) The accumulation. Endocrinology, 120r26P271. origin of citrulline-containing proteins in the hair follicle and the Bodner, M., Castillo, J.-L., Theil, L.E., Deerinck, T., Ellisman, M., and chemical nature of trichohyalin, an intracellular precursor. BioKarin, M. (1988)The pituitary-specific transcription factor GHF-1 chim. Biophys. Acta, 495:159-175. is a homeobox-containingprotein. Cell, 55:505-518. C., and Spelsberg, T.C. (1989) Ovarian steroid action on gene Borrelli, E., Heyman, R.A., Arias, C., Sawchenko, P.E., and Evans, Rories, expression: Mechanisms and models. Annu. Rev. Physiol., 52:653R.M. (1989) Transgenic mice with inducible dwarfism. Nature, 681. 339.538441. Butcher, R.L., Collins, W.E., and Fugo, N.W. (1974) Plasma concen- Rothnagel, J.A., and Rogers, G.E. (1984) Citrulline in proteins from the enzymatic deimination of arginine residues. In: Methods of tration of LH, FSH, prolactin, progesterone and estradiol-17 Enzymology. F. Wold and K. Moldave, eds. Academic Press, Inc. throughout the 4 day estrous cycle of the rat. Endocrinology, Orlando, FL, Vol. 107, pp. 624-631. 94:1704-1711. Dolle, P., Castillo, J.-L., Theil, L.E., Deerinck, Ellisman, M., and Senshu, T., Akiyama, K., Nagata, S., Watanabe, K., and Hikichi, K. (1989) Peptidylarginine deiminase in rat pituitary: Sex difference, Karin, M. (1990)Expression of GHF-1 protein in mouse pituitaries estrous cycle-related changes, and estrogen dependence. Endocricorrelates both temporally and spatially with the onset of growth nology, 224:2666-2670. hormone gene activity. Cell, 60:809-820. Fujisaki, S., and Sugawara, K. (1981) Properties of peptidylarginine Takahara, H., Oikawa, Y., and Sugaware, K. (1983)Purification and characterization of peptidylarginine deiminase from rabbit skeletal deiminase from the epidermis of newborn rat. J. Biochem. (Tokyo), muscle. J. Biochem. (Tokyo),94:1945-1963. 89:257-263. H., Okamoto, H., and Sugawara, K. (1985) Specific modiInagaki, M., Takahara, H., Nishi, Y., and Sugawara, K. (1989) Takahara, fication of the functional arginine residue in soybean trypsin Ca2+-dependentdeimination-induced disassembly of intermediate inhibitor (Kunitz) by peptidylarginine deiminase. J. Biol. Chem., filaments involves sDecific modification of the amino-terminal head ~ ~ ~ . . ~ ~ ~~. 260:837-383. domain. J. Biol. Chim. 264:18119-18127. Ingraham, H.A., Chen, R., Mangalam, H.J., Elisholtz, H.P., Flynn, Takahara, H., Tsuchida, M., Kusubata, S., Akutsu, K., Tagami, S., and Sugawara, K. (1989)Peptidylarginine deiminase of the mouse: S.E.. Lin. C.R., Simmons. D.M.. Swanson. L.. and Rosenfeld. M.G. Distribution, properties and immunocytochemical localization. J. (1988) A’ tissue-specific transcription factor containing a homeBiol. Chem., 264:13361-13368. odomain specifies a pituitary phenotype. Cell, 55:519-529. Inoue, K., Hattori, M., Sakai, T., Inukai, S., and Ito, A. (1990) Vician, L., Shupnik, M.A., and Gorski, J. (1979)Effects ofestrogen on primary ovine pituitary cell cultures: Stimulation of prolactin Establishment of a series of pituitary clonal cell lines differing in secretion, synthesis, and preprolactin messenger ribonucreic acid morphology and response to estrogen. Endocrinology, 2262313activity. Endocrinology, 104:736743. 2320. Ito, A., Kawashima, K., and Fujimoto, N., Watanabe, H., and Naito, Watanabe, K., Akiyama, A., Hikichi, K., Ohtsuka, R., Okuyama, A,, and Senshu, T. (1988)Combined biochemical and immunochemical M. (1985)Inhibition by 2-bromo- -ergocriptine and tamoxifen of the comparison of peptidylarginine deiminases present in various tisgrowth of an estrogen-dependent transplantable pituitary tumor sues. Biochim. Biophys. Acta, 966:375-383. (MtTiF84)in F344 rats. Cancer Res., 456436-6441. Kassis, J.A., Sakai, D., Walent, J.H., and Gorski, J. (1984) Primary Westley, B., and Rochefort, H. (1980)A secreted glycoprotein induced in human breast cancer cell lines. Cell, 20:353-362. cultures of estrogen-responsive cells from rat uteri: induction of progesterone receptors and a secreted protein. Endocrinology, Wood, D.D., and Moscarello, M.A. (1989) The isolation, characterization, and lipid-aggregating properties of a citrulline containing 224:155&1566. myelin basic protein. J. Biol. Chem., 264.5121-5127. Kubilus, J., and Baden, H.P. (1983) Purification and properties of a brain enzyme which deiminates proteins. Biochim. Biophys. Acta, 745:285-291. ~

~

~

~~

~~

~

~~