In VitroCell. Dev. Biol.26:431-440. May 1990 9 1990TissueCulture Association 0883-8364/90 $01.50+0.00
GONADOTROPES IN A NOVEL RAT PITUITARY TUMOR CELL LINE, RC-4B/C. ESTABLISHMENT AND PARTIAL CHARACTERIZATION OF THE CELL LINE ILSE HURBAIN-KOSMATH, ANNETTE BERAULT, NADINE NOEL, JOLANTA POLKOWSKA, ANNE BOHIN, MARIAN JUTISZ', EDWARD H. LEITER, WESLEY G. BEAMER, HENDRICK G. BEDIGIAN, MURIEL T. DAVISSON, ANDDAVID E. HARRISON Laboratoire des Hormones Polypeptidiques, CNRS, 91198 Gif sur yvette, France (I. H-K., A. B., N. N., J. P., A. Bo., M. J.), and The Jackson Laboratory, Bar Harbor, Maine 04609 (E. H. L., IV. G. B., H. G. B., M. T. D., D. E. H.)
tReceived 13 September 1989; accepted 12 December 1989)
SUMMARY An epithelial cell line (RC-4B/C) was established from a pituitary, adenoma obtained from a 3-yr-old ( A C I / f M a i ) < F344/fMai)F1 male rat. Before Year 5 in vitro, RC-4B/C cells could not be viably recovered from cryogenic storage. RecoveD" of viable cells from cryogenic storage in Year 5 was associated with a more transformed phenotype, including the appearance of endogenous C-type rat retroviral particles. The ultrastructural appearance of the cells was similar to that of differentiated anterior pituitary cells; the cultured cells contained numerous, electron dense, secretory granules, Golgi complexes, and extended arrays of rough endoplasmic reticulum. Immunocytochemical study showed that all cell types present in the rat anterior pituitary gland were present in the cell line. The percentage of luteinizing hormone beta (LH/31 cells in the cell line was higher (19.9%) and that of growth hormone cells was lower (12.2%) than in normal male rat pituitary, whereas the cell line contained a comparable percentage of follicle stimulating hormone beta (FSH/~), prolactin (PRLI, ACTH, and thyrotropin beta cells. Radioimmunoassay data demonstrated the P R L content of the cells was comparable to that of normal male rat pituitary gland, whereas the content of LH and F S H was 70- and 800-fold lower, respectively. Assay of specific receptor sites for gonadotropin releasing hormone (GnRH) using Scatchard plots of the data established the RC-4B/C cells contained GnRH receptor sites of the same affinity as in the pituitary, gland, but of twofold lower capacity. These data suggest the RC4 B / C cell line warrants further study as a model for the induction and maintenance of the gonadotropic function of the pituitary gland. K e y words: gonadotropes; pituitary adenoma; cell line; anterior pituitary cells; G n R H receptor. Although pituitary gonadotrope adenomas have been described in aged rats of certain strains I3,27,28), very little functional characterization of the tumor cells has been done. Shiino et al. (31) described a gonadotrope-rich cell line established from cloned pituitary cells (2A8) implanted under the kidney capsule of hypophysectomized female rats. However, electron microscopy showed these cells to be agranular, and a permanent cell line was not established. In this report we describe the establishment of a permanent epithelial cell line, RC-4B/C, from an aged rat pituitary adenoma and its culture in a well-defined medium containing a semisynthetic substitute. We use immunocytochemical techniques to demonstrate that the ultrastructurally well-differentiated cells continue to produce, albeit at altered ratios, the spectrum of polypeptide hormones found in the anterior pituitary. This cell line seems to contain gonadotropes, although gonadotropin content of the cells is much lower than in the pituitary cells from normal male rat. The presence of gonadotropin releasing hormone (GnRH) receptors on RC-4B/C cells warrants their use for regulation or transfection studies or both.
| NTRODUCTION Hyperseeretion of the glyeoprotein hormones luteinizing hormone (LH), follicle stimulating hormone (FSH), and thyrotropin (TSH) from human pituitary, adenomas is uncommon {for review, see (21,32)]. When cases where adenorna-derived excess gonadotropin are encountered, FSH is typically involved; concomitant LH secretion is variable, and secretion of a and FSH/3 subunits is abnormally governed (18,32,33,41). Data from pituitary function tests indicate that pituitary adenomas are often not under hypothalamic control, thus complicating the physician's task of identifying the abnormality present and the treatment indicated. Spontaneously occurring or induced adenomas of the anterior pituitary gland in rodents have permitted the establishment of permanent cell lines that continue to produce and secrete various peptide hormones, such as growth hormone {GH) (35), prolactin (PRL) (36), and ACTH (4). Permanent cell lines that secrete one or both gonadotropins, FSH and LH, have proven more elusive. To whom correspondence should be addressed. 431
432
HURBAIN-KOSMATH ET AL.
MATERIALS AND METHODS
Establishment of the RC-4B/C cell line. A pituitary adenoma 6 to 7 mm in diameter from a 3-yr-old ( A C I / f M a i ) < F344/fMai)F1 male rat, discovered in 1977 in an aging colony maintained at The Jackson Laboratory by one of us (D.E.H.), was aseptically excised and diced into small pieces with scalpels. The minced tissue was enzymatically dissociated into a single cell suspension by the method of Vale et al. ~40), with the modification that a reciprocating-shaker water bath was used in place of an electric stirrer. The cell suspension produced was inoculated into three siliconized 25-ml Erlenmeyer flasks at a density of 1.25 X l0 s cells-3 ml -I culture medium" flask. Gyrorotatory shaking of these suspensions at 37 ~ C and 84 rpm in a water bath shaker (New Brunswick Scientific, New Brunswick, N J, model G76~ produced aggregates which were harvested at Day 6 of culture and inoculated into 25-cm 2 plastic culture flasks (Lux Scientific, Newbury Park, CAL The cell aggregates, formed after 6 d of gyrorotatory shaking of the dispersed pituitary tumor cell suspension, attached within 24 h after plating in 25-cm 2 culture flasks. Clumps of attached cells were primarily epitheliold in morphology. Outgrowth of fibroblastoid cells from the epithelioid cell clusters occurred within 6 d after plating; the growth of these fibroblastoid cells was halted and the fibroblasts slowly eliminated during an ensuing 3-mo. period by serum deprivation for 12-d intervals followed by serum restoration for 3-d intervals. In one flask, epidermal growth factor (EGF, 2.5 ng/ml) was included in both the serum-free and serum-containing medium changes after Day 41 of culture. Subsequently, epithelioid cells in this flask began to proliferate, whereas fibroblastoid cells decreased in number. The first subculture was performed at culture Day 96; subsequent serial passage at 3- to 4-wk intervals over a period of 1 yr thereafter allowed selection of a fibroblast-free subline designated RC-4B/C. At The Jackson Laboratory where the cell line was established ~by E.H.L.), the starting culture medium was Dulbecco's modified Eagle's minimum essential medium ( D M E M ) containing 16.5 m M glucose, 15 m M N-2hydroxyethyl-piperazine-N-2-ethanesuifonic acid tHEPES), 50 ~g/ml gentamicin, 0.2 m g / m l bovine serum albumin IBSAL Eagle's nonessential amino acid supplement, and 10% dialyzed, heat-inactivated fetal bovine serum ~FBS) tall tissue culture media and FBS were obtained from GIBCO, Grand Island, NY, and all chemicals from Sigma Chemical Co., St. Lofiis, MO). After the cell line was established, a 1:1 mixture of D M E M and Eagle's minimum essential medium alpha (MEM-alpha) ithe latter containing riboand deoxyribonucleosides) plus 2.5 ng/ml E G F (Collaborative Research, Lexington, MA), as well as the D M E M supplements enumerated above, was used routinely as a maintenance medium at The Jackson Laboratory. Cultures were maintained at 34 ~ C and the medium was changed twice weekly. Cells were subcuhured at 2-wk intervals. A Mycoplasma arginini contaminant detected during Year 2 of culture was successfully eliminated by combined treatment with three antibiotics (gentamicin sulfate, tetracycline, and chloramphenicol) as described by Ulrich and Briand {39).
Culture of the RC-4B/C cells. In 1986, Passage 268 cells were brought to Gif sur Yvette, where FBS was progressively replaced with a semisynthetic serum supplement, Nu-Serum (Collaborative Research, Lexington, MA). All immunocytochemical studies were performed on RC-4B/C cells cultured in the medium containing 5% Nu-Serum in place of FBS. According to the supplier (personal communication), Nu-Serum contains among other components, 25~ of steroid-frce newborn bovine serum, 3.40 ng/ml 17p-estradiol, 4.25 ng/ml progesterone, and 3.86 ng/ml testosterone. The cell line is currently maintained by plating into 175-cm 2 plastic tissue culture flasks tNunc, Roskilde, Denmark) at a density of 3 )< los cells/30 ml and gassed with a mixture 10% CO2:20% 05:70% N2. Cultures were maintained at 34 ~ C, and the medium was changed twice weekly. Cells were passagecl at 2-wk intervals. Culture medium was replaced by 8 ml of 0.005~ trypsin (GIBCO) in saline containing 0.02% E D T A and 0.1% glucose. Mter a few minutes, cells were centrifuged for 2 min at 450 g and suspended in fresh culture medium. For cryogenic storage, 5 )< 106 cells/ml were placed in medium containing 10% dimethyl suffoxide and 10% Nu~ kept at - - 2 0 ~ C for 4 h, --40 ~ C overnight, --80 ~ C for 7 to 8 h, and then were transferred to liquid nitrogen. For immunocytochemical studies, cells were cultured in 9-cm 2 plastic slide flasks (Nunc). During the past 2 yr the cell line was screened at least every 3 mo. for mycoplasmal contamination by staining with Hoechst 33258 (Hoechst, Franldurth, FRG) fluorescent dye, which binds specifically to D N A t6). Another mycoplasmal contamination detected at the end of 1988 was eliminated as previously described 139). Growth curve. Growth rate of the RC-4B/C cell line was established as follows. Cells were cultured for 7 passages after recovery from cryogenic storage, then cells were plated into 9.6-cm 2 plastic tissue culture dishes (Nunc) at a density of l 0 s cells/3 ml and gassed as above. Cultures were maintained at 34 ~ C and the medium was changed twice weekly. During a period of 16 d, three dishes were removed each day at noon; the medium was discarded, the cells were frozen at --80 ~ C and D N A was assayed in the ceils according to the method of Labarca and Paigen ~22). Cells were counted in three other dishes by means of a Coulter counter. Culture of normal male rat anterior pituitary cells. Primary cultures established from normal male rat pituitary cells were immunocytochemically analyzed as a reference for comparison of distribution frequencies of differentiated pituitary cells with those observed in the R C - 4 B / C cell line. Primary cultures of pituitary cells were also used as a reference for LH, FSH, and P R L radioimmunoassays IRIAs) and for G n R H radioreeeptor assay. Adult male rat (Wistarj pituitary glands were enzymatically dispersed by trypsin digestion using the method of Hopkins and Farquhar t15) as modified by Denef et al. (10L Pituitary cells were cultured for 3 d in H a m ' s F10 medium IEurobio, Paris, France) supplemented with 10% FBS, 25 m M H E P E S , and antibiotics. Chromosome and retrovirus analysis. To obtain chromosome preparations, cells were preincubated in 5
GONADOTROPES IN A NOVEL PITUITARY CELL LINE gg/ml colchicine for 30 min; then mitotic cells were detached from culture flasks by shaking, transferred to hypotonic saline (0.75 M KC1) for 15 min, and fixed with methanol:acetic acid (3:lL Air-dried preparations were stained with 2% Gurr Giemsa in Gurr phosphate buffer ipH 6.8) for chromosome counting. Modal chromosome numbers were determined from Giemsa-banded chromosome preparations. The staining method and photography methods have been described previously (9). To identify a budding retrovirus observed ultrastructurally in the RC-4B/C rat cell line, supernatant fluids (15 to 20 ml) from the cultures were concentrated by ultracentrifugation and assayed for manganesepreferring RNA-directed D N A polymerase ( R D D P ) activity as described (23}. To determine if the virus expressed in the RC-4B/C cell line was of rat origin or represented a contaminating mouse retrovirus (eeotropic or xenotropic), in vitro host range studies were done. Single cell suspensions of RC-4B/C cells were treated with mitomyein C (25 /~g for 30 rain) and 1 X 10~ cells cocultivated with SC-1 mouse fibroblasts (which support the growth of ecotropic murine leukemia virus, MuLV), or with a mink lung cell line (which supports xenotropic MuLV growth} (2). After 28 d in culture, supernatant fluids from the cocultivated cultures were assayed for R D D P activity. Inasmuch as ecotropic MuLVs are able to infect rat cells at a low efficiency, the UV-XC plaque assay (2) was also used to establish whether an ecotropic MuLV was present in RC-4B/C cells. Supernatant fluids from RC-4B/C cell cultures or mitomycin C-treated RC-4B/C cells were placed onto SC-1 mouse fibroblast cells. Individual SC-1 cell cultures were exposed to ultraviolet radiation and at either 5, 14, or 28 d postiufection, co-cultivated with XC cells as described elsewhere (2). Finally, the RC-4B/C cells were screened for the presence of somatically acquired ecotropic MuLV DNA by the whole cell blot technique of Joyner et al. (19), using the 400 base pair ecotropic MuLV-specific D N A probe (5). Morphologic and immunocytochemical characterization. At the time of primary tumor extirpation, 1- to 2-ram sized fragments were fixed in Bouin's fluid, dehydrated, embedded in paraffin, and 5-gm stained sections were examined by light microscopy. Samples of the cultured cell line were prepared for ultrastructural analysis by fixation in 3% glutaraldehyde1% formaldehyde in 0.1 M sodium eacodylate buffer, pH 7.4, and embedded in Epon-Araldite resin. Ultrathin sections were examined with a Hitachi HU-11C transmission microscope. For immunocytoehemistry, RC-4B/C cells passaged between 10 and 15 times after recovery from cryogenic storage were cultured at 34~ C for 3 d in plastic slide flasks. Cells were washed rapidly with 0.1 M phosphate buffered saline (PBS), pH 7.4, and then fixed for 3 h at room temperature (RT) in paraformaldehyde-picric acid fixative buffered with 0.1 M PBS, pH 7.4 (42). Cells were washed 4 times over a period of 2 h in 0.01 M PBS and once in 0.01 M PBS containing 1% normal ovine serum and then stained by the peroxidase-labeled antibody method (29L Sequential incubations with antisera were
433
done with the following solutions for the times indicated: primary antiserum in 0.01 M PBS, 24 to 72 h at 4 ~ C; 1:40 goat antirabbit y-globulin bound to peroxidase (A-IgGPer, Institut Pasteur, Paris, France), 1 h at RT. All antibodies were diluted in 0.01 M PBS containing 1% normal ovine serum. The cells were washed 3 times with 0.01 M PBS after exposure to the primary antisera and A-IgG-Per. Peroxidase localization was revealed with a freshly prepared 3-3' -diamino-benzidine tetrahydrochloride (DAB) (Sigma) containing 0.005% of hydrogen peroxide (14L After the DAB treatment, cells were additionally stained by the silver intensification method of Liposits et al. (25L Antisera. Anti-ovine LHfl (oLHfl) (I3) and anti-ovine FSHfl (oFSHfl) (P3) sera as well as ovine gonadotropin subunits LHa, LHf3, FSHfl, were prepared in our laboratory (Gif sur YvetteL The specificity of antisera was tested using R I A methods. At the dilutions used, the following cross-reactivities were observed: anti-oLHfl did not cross-react with ovine LI-Ia (oLHa) or rat P R L (rPRL) and cross-reacted < 2% with rat FSH (rFSH); anti-oFSHfl did not cross-react with either oLI-Ia, rLH, or rPRL. Absorption tests of anti-oLHfl showed no cross-reactivity with r P R L and only very slight reactivity with oFSHfl and rFSH; anti-oFSHfl did not cross-react with either oLHfl, rLH, or rPRL. Absorption tests of antiporcine LHfl (pLHfl) (no. 19526) showed no cross-reactivity with either rFSH, oFSHfl, or rPRL; anti-rPRL (no. 19601) did not cross-react with either rLH, rFSH, oLHfl, or oFSHfl. AntipLHfl (no. 19526), anti-rPRL (no. 19601L antiporcine TSHf3 (pTSHfl) (no. 1238L anti-ACTH, 2, (no. 7.27.02.70), antiequine GH (eGH) (no. 19538) were donated by Dr. M.P. Dubois (INRA, Nouzilly, FranceL The specificity controls of the last two antisera were described previously (11,12L Rat pituitary hormones used as antigens, rLH32V02, rFSH32V02, rPRL32V03, were provided through the National Hormone and Pituitary Program. Antisera were used at following dilutions for immunocytochemistry: Anti-oLHfl, 1:5000 to 1:10 000; anti-pLH/L 1:2000 to 1:5000; anti-oFSHfl, 1:1000 to 1:2000; anti-rPRL, 1:3000 to 1:5000; anti-pTSHfL 1:2000 to 1:5000; antiACTH~.2,, 1:2000 to 1:5000; anti-eGH, 1:5000 to 1:10 000. Immunocytochemical controls. To establish specificity of the staining, one of the major components in the mixture was omitted during the staining procedure. This consisted of using PBS buffer with 1% BSA or 1% normal rabbit serum (Biosys, Compiegne, France} instead of an antiserum, or using normal sheep serum (Institut Pasteur, Paris, France} instead of AIgG-Per. To test specificity of the staining, antisera were incubated with increasing concentrations of various hormones or hormone subunits for 48 h at 4 ~ C. The resulting solutions were then used for staining dispersed pituitary cells. Cell type quantitation. Once stained, the slides were quantitatively analyzed for presence of various cell types. This was done by counting positively stained cells by means of a grid mounted in the ocular lens of the microscope. Usually three slides stained with the same antiserum were chosen. The area of a slide for counting was determined in a random fashion, and a given cell type was counted in an area containing a total population of 800 to 2000 cells. For the cell counting we considered that the cell was positively stained when its cytoplasm was
434
HURBAIN-KOSMATH ET AL.
totally or partly stained. Cells with only peripheral staining were not counted. The black grains observed in
some unstained cells represent silver grains not completely removed in the silver intensification method ~25).
FIG. 1. Morphology of RC-4B/C rat pituitary tumor cell line. a, Light micrograph of 5-/am section of donor ACI/fMai pituitary adenoma. Mitotic figures are present (arrows~. X427. b, Phase microscopy of the RC-4B/C cell line showing the epithelioid nature of the cells. )
GONADOTROPES IN A NOVEL RAT PITUITARY TUMOR CELL LINE, RC-4B/C. ESTABLISHMENT AND PARTIAL CHARACTERIZATION OF THE CELL LINE ILSE HURBAIN-KOSMATH, ANNETTE BERAULT, NADINE NOEL, JOLANTA POLKOWSKA, ANNE BOHIN, MARIAN JUTISZ', EDWARD H. LEITER, WESLEY G. BEAMER, HENDRICK G. BEDIGIAN, MURIEL T. DAVISSON, ANDDAVID E. HARRISON Laboratoire des Hormones Polypeptidiques, CNRS, 91198 Gif sur yvette, France (I. H-K., A. B., N. N., J. P., A. Bo., M. J.), and The Jackson Laboratory, Bar Harbor, Maine 04609 (E. H. L., IV. G. B., H. G. B., M. T. D., D. E. H.)
tReceived 13 September 1989; accepted 12 December 1989)
SUMMARY An epithelial cell line (RC-4B/C) was established from a pituitary, adenoma obtained from a 3-yr-old ( A C I / f M a i ) < F344/fMai)F1 male rat. Before Year 5 in vitro, RC-4B/C cells could not be viably recovered from cryogenic storage. RecoveD" of viable cells from cryogenic storage in Year 5 was associated with a more transformed phenotype, including the appearance of endogenous C-type rat retroviral particles. The ultrastructural appearance of the cells was similar to that of differentiated anterior pituitary cells; the cultured cells contained numerous, electron dense, secretory granules, Golgi complexes, and extended arrays of rough endoplasmic reticulum. Immunocytochemical study showed that all cell types present in the rat anterior pituitary gland were present in the cell line. The percentage of luteinizing hormone beta (LH/31 cells in the cell line was higher (19.9%) and that of growth hormone cells was lower (12.2%) than in normal male rat pituitary, whereas the cell line contained a comparable percentage of follicle stimulating hormone beta (FSH/~), prolactin (PRLI, ACTH, and thyrotropin beta cells. Radioimmunoassay data demonstrated the P R L content of the cells was comparable to that of normal male rat pituitary gland, whereas the content of LH and F S H was 70- and 800-fold lower, respectively. Assay of specific receptor sites for gonadotropin releasing hormone (GnRH) using Scatchard plots of the data established the RC-4B/C cells contained GnRH receptor sites of the same affinity as in the pituitary, gland, but of twofold lower capacity. These data suggest the RC4 B / C cell line warrants further study as a model for the induction and maintenance of the gonadotropic function of the pituitary gland. K e y words: gonadotropes; pituitary adenoma; cell line; anterior pituitary cells; G n R H receptor. Although pituitary gonadotrope adenomas have been described in aged rats of certain strains I3,27,28), very little functional characterization of the tumor cells has been done. Shiino et al. (31) described a gonadotrope-rich cell line established from cloned pituitary cells (2A8) implanted under the kidney capsule of hypophysectomized female rats. However, electron microscopy showed these cells to be agranular, and a permanent cell line was not established. In this report we describe the establishment of a permanent epithelial cell line, RC-4B/C, from an aged rat pituitary adenoma and its culture in a well-defined medium containing a semisynthetic substitute. We use immunocytochemical techniques to demonstrate that the ultrastructurally well-differentiated cells continue to produce, albeit at altered ratios, the spectrum of polypeptide hormones found in the anterior pituitary. This cell line seems to contain gonadotropes, although gonadotropin content of the cells is much lower than in the pituitary cells from normal male rat. The presence of gonadotropin releasing hormone (GnRH) receptors on RC-4B/C cells warrants their use for regulation or transfection studies or both.
| NTRODUCTION Hyperseeretion of the glyeoprotein hormones luteinizing hormone (LH), follicle stimulating hormone (FSH), and thyrotropin (TSH) from human pituitary, adenomas is uncommon {for review, see (21,32)]. When cases where adenorna-derived excess gonadotropin are encountered, FSH is typically involved; concomitant LH secretion is variable, and secretion of a and FSH/3 subunits is abnormally governed (18,32,33,41). Data from pituitary function tests indicate that pituitary adenomas are often not under hypothalamic control, thus complicating the physician's task of identifying the abnormality present and the treatment indicated. Spontaneously occurring or induced adenomas of the anterior pituitary gland in rodents have permitted the establishment of permanent cell lines that continue to produce and secrete various peptide hormones, such as growth hormone {GH) (35), prolactin (PRL) (36), and ACTH (4). Permanent cell lines that secrete one or both gonadotropins, FSH and LH, have proven more elusive. To whom correspondence should be addressed. 431
432
HURBAIN-KOSMATH ET AL.
MATERIALS AND METHODS
Establishment of the RC-4B/C cell line. A pituitary adenoma 6 to 7 mm in diameter from a 3-yr-old ( A C I / f M a i ) < F344/fMai)F1 male rat, discovered in 1977 in an aging colony maintained at The Jackson Laboratory by one of us (D.E.H.), was aseptically excised and diced into small pieces with scalpels. The minced tissue was enzymatically dissociated into a single cell suspension by the method of Vale et al. ~40), with the modification that a reciprocating-shaker water bath was used in place of an electric stirrer. The cell suspension produced was inoculated into three siliconized 25-ml Erlenmeyer flasks at a density of 1.25 X l0 s cells-3 ml -I culture medium" flask. Gyrorotatory shaking of these suspensions at 37 ~ C and 84 rpm in a water bath shaker (New Brunswick Scientific, New Brunswick, N J, model G76~ produced aggregates which were harvested at Day 6 of culture and inoculated into 25-cm 2 plastic culture flasks (Lux Scientific, Newbury Park, CAL The cell aggregates, formed after 6 d of gyrorotatory shaking of the dispersed pituitary tumor cell suspension, attached within 24 h after plating in 25-cm 2 culture flasks. Clumps of attached cells were primarily epitheliold in morphology. Outgrowth of fibroblastoid cells from the epithelioid cell clusters occurred within 6 d after plating; the growth of these fibroblastoid cells was halted and the fibroblasts slowly eliminated during an ensuing 3-mo. period by serum deprivation for 12-d intervals followed by serum restoration for 3-d intervals. In one flask, epidermal growth factor (EGF, 2.5 ng/ml) was included in both the serum-free and serum-containing medium changes after Day 41 of culture. Subsequently, epithelioid cells in this flask began to proliferate, whereas fibroblastoid cells decreased in number. The first subculture was performed at culture Day 96; subsequent serial passage at 3- to 4-wk intervals over a period of 1 yr thereafter allowed selection of a fibroblast-free subline designated RC-4B/C. At The Jackson Laboratory where the cell line was established ~by E.H.L.), the starting culture medium was Dulbecco's modified Eagle's minimum essential medium ( D M E M ) containing 16.5 m M glucose, 15 m M N-2hydroxyethyl-piperazine-N-2-ethanesuifonic acid tHEPES), 50 ~g/ml gentamicin, 0.2 m g / m l bovine serum albumin IBSAL Eagle's nonessential amino acid supplement, and 10% dialyzed, heat-inactivated fetal bovine serum ~FBS) tall tissue culture media and FBS were obtained from GIBCO, Grand Island, NY, and all chemicals from Sigma Chemical Co., St. Lofiis, MO). After the cell line was established, a 1:1 mixture of D M E M and Eagle's minimum essential medium alpha (MEM-alpha) ithe latter containing riboand deoxyribonucleosides) plus 2.5 ng/ml E G F (Collaborative Research, Lexington, MA), as well as the D M E M supplements enumerated above, was used routinely as a maintenance medium at The Jackson Laboratory. Cultures were maintained at 34 ~ C and the medium was changed twice weekly. Cells were subcuhured at 2-wk intervals. A Mycoplasma arginini contaminant detected during Year 2 of culture was successfully eliminated by combined treatment with three antibiotics (gentamicin sulfate, tetracycline, and chloramphenicol) as described by Ulrich and Briand {39).
Culture of the RC-4B/C cells. In 1986, Passage 268 cells were brought to Gif sur Yvette, where FBS was progressively replaced with a semisynthetic serum supplement, Nu-Serum (Collaborative Research, Lexington, MA). All immunocytochemical studies were performed on RC-4B/C cells cultured in the medium containing 5% Nu-Serum in place of FBS. According to the supplier (personal communication), Nu-Serum contains among other components, 25~ of steroid-frce newborn bovine serum, 3.40 ng/ml 17p-estradiol, 4.25 ng/ml progesterone, and 3.86 ng/ml testosterone. The cell line is currently maintained by plating into 175-cm 2 plastic tissue culture flasks tNunc, Roskilde, Denmark) at a density of 3 )< los cells/30 ml and gassed with a mixture 10% CO2:20% 05:70% N2. Cultures were maintained at 34 ~ C, and the medium was changed twice weekly. Cells were passagecl at 2-wk intervals. Culture medium was replaced by 8 ml of 0.005~ trypsin (GIBCO) in saline containing 0.02% E D T A and 0.1% glucose. Mter a few minutes, cells were centrifuged for 2 min at 450 g and suspended in fresh culture medium. For cryogenic storage, 5 )< 106 cells/ml were placed in medium containing 10% dimethyl suffoxide and 10% Nu~ kept at - - 2 0 ~ C for 4 h, --40 ~ C overnight, --80 ~ C for 7 to 8 h, and then were transferred to liquid nitrogen. For immunocytochemical studies, cells were cultured in 9-cm 2 plastic slide flasks (Nunc). During the past 2 yr the cell line was screened at least every 3 mo. for mycoplasmal contamination by staining with Hoechst 33258 (Hoechst, Franldurth, FRG) fluorescent dye, which binds specifically to D N A t6). Another mycoplasmal contamination detected at the end of 1988 was eliminated as previously described 139). Growth curve. Growth rate of the RC-4B/C cell line was established as follows. Cells were cultured for 7 passages after recovery from cryogenic storage, then cells were plated into 9.6-cm 2 plastic tissue culture dishes (Nunc) at a density of l 0 s cells/3 ml and gassed as above. Cultures were maintained at 34 ~ C and the medium was changed twice weekly. During a period of 16 d, three dishes were removed each day at noon; the medium was discarded, the cells were frozen at --80 ~ C and D N A was assayed in the ceils according to the method of Labarca and Paigen ~22). Cells were counted in three other dishes by means of a Coulter counter. Culture of normal male rat anterior pituitary cells. Primary cultures established from normal male rat pituitary cells were immunocytochemically analyzed as a reference for comparison of distribution frequencies of differentiated pituitary cells with those observed in the R C - 4 B / C cell line. Primary cultures of pituitary cells were also used as a reference for LH, FSH, and P R L radioimmunoassays IRIAs) and for G n R H radioreeeptor assay. Adult male rat (Wistarj pituitary glands were enzymatically dispersed by trypsin digestion using the method of Hopkins and Farquhar t15) as modified by Denef et al. (10L Pituitary cells were cultured for 3 d in H a m ' s F10 medium IEurobio, Paris, France) supplemented with 10% FBS, 25 m M H E P E S , and antibiotics. Chromosome and retrovirus analysis. To obtain chromosome preparations, cells were preincubated in 5
GONADOTROPES IN A NOVEL PITUITARY CELL LINE gg/ml colchicine for 30 min; then mitotic cells were detached from culture flasks by shaking, transferred to hypotonic saline (0.75 M KC1) for 15 min, and fixed with methanol:acetic acid (3:lL Air-dried preparations were stained with 2% Gurr Giemsa in Gurr phosphate buffer ipH 6.8) for chromosome counting. Modal chromosome numbers were determined from Giemsa-banded chromosome preparations. The staining method and photography methods have been described previously (9). To identify a budding retrovirus observed ultrastructurally in the RC-4B/C rat cell line, supernatant fluids (15 to 20 ml) from the cultures were concentrated by ultracentrifugation and assayed for manganesepreferring RNA-directed D N A polymerase ( R D D P ) activity as described (23}. To determine if the virus expressed in the RC-4B/C cell line was of rat origin or represented a contaminating mouse retrovirus (eeotropic or xenotropic), in vitro host range studies were done. Single cell suspensions of RC-4B/C cells were treated with mitomyein C (25 /~g for 30 rain) and 1 X 10~ cells cocultivated with SC-1 mouse fibroblasts (which support the growth of ecotropic murine leukemia virus, MuLV), or with a mink lung cell line (which supports xenotropic MuLV growth} (2). After 28 d in culture, supernatant fluids from the cocultivated cultures were assayed for R D D P activity. Inasmuch as ecotropic MuLVs are able to infect rat cells at a low efficiency, the UV-XC plaque assay (2) was also used to establish whether an ecotropic MuLV was present in RC-4B/C cells. Supernatant fluids from RC-4B/C cell cultures or mitomycin C-treated RC-4B/C cells were placed onto SC-1 mouse fibroblast cells. Individual SC-1 cell cultures were exposed to ultraviolet radiation and at either 5, 14, or 28 d postiufection, co-cultivated with XC cells as described elsewhere (2). Finally, the RC-4B/C cells were screened for the presence of somatically acquired ecotropic MuLV DNA by the whole cell blot technique of Joyner et al. (19), using the 400 base pair ecotropic MuLV-specific D N A probe (5). Morphologic and immunocytochemical characterization. At the time of primary tumor extirpation, 1- to 2-ram sized fragments were fixed in Bouin's fluid, dehydrated, embedded in paraffin, and 5-gm stained sections were examined by light microscopy. Samples of the cultured cell line were prepared for ultrastructural analysis by fixation in 3% glutaraldehyde1% formaldehyde in 0.1 M sodium eacodylate buffer, pH 7.4, and embedded in Epon-Araldite resin. Ultrathin sections were examined with a Hitachi HU-11C transmission microscope. For immunocytoehemistry, RC-4B/C cells passaged between 10 and 15 times after recovery from cryogenic storage were cultured at 34~ C for 3 d in plastic slide flasks. Cells were washed rapidly with 0.1 M phosphate buffered saline (PBS), pH 7.4, and then fixed for 3 h at room temperature (RT) in paraformaldehyde-picric acid fixative buffered with 0.1 M PBS, pH 7.4 (42). Cells were washed 4 times over a period of 2 h in 0.01 M PBS and once in 0.01 M PBS containing 1% normal ovine serum and then stained by the peroxidase-labeled antibody method (29L Sequential incubations with antisera were
433
done with the following solutions for the times indicated: primary antiserum in 0.01 M PBS, 24 to 72 h at 4 ~ C; 1:40 goat antirabbit y-globulin bound to peroxidase (A-IgGPer, Institut Pasteur, Paris, France), 1 h at RT. All antibodies were diluted in 0.01 M PBS containing 1% normal ovine serum. The cells were washed 3 times with 0.01 M PBS after exposure to the primary antisera and A-IgG-Per. Peroxidase localization was revealed with a freshly prepared 3-3' -diamino-benzidine tetrahydrochloride (DAB) (Sigma) containing 0.005% of hydrogen peroxide (14L After the DAB treatment, cells were additionally stained by the silver intensification method of Liposits et al. (25L Antisera. Anti-ovine LHfl (oLHfl) (I3) and anti-ovine FSHfl (oFSHfl) (P3) sera as well as ovine gonadotropin subunits LHa, LHf3, FSHfl, were prepared in our laboratory (Gif sur YvetteL The specificity of antisera was tested using R I A methods. At the dilutions used, the following cross-reactivities were observed: anti-oLHfl did not cross-react with ovine LI-Ia (oLHa) or rat P R L (rPRL) and cross-reacted < 2% with rat FSH (rFSH); anti-oFSHfl did not cross-react with either oLI-Ia, rLH, or rPRL. Absorption tests of anti-oLHfl showed no cross-reactivity with r P R L and only very slight reactivity with oFSHfl and rFSH; anti-oFSHfl did not cross-react with either oLHfl, rLH, or rPRL. Absorption tests of antiporcine LHfl (pLHfl) (no. 19526) showed no cross-reactivity with either rFSH, oFSHfl, or rPRL; anti-rPRL (no. 19601) did not cross-react with either rLH, rFSH, oLHfl, or oFSHfl. AntipLHfl (no. 19526), anti-rPRL (no. 19601L antiporcine TSHf3 (pTSHfl) (no. 1238L anti-ACTH, 2, (no. 7.27.02.70), antiequine GH (eGH) (no. 19538) were donated by Dr. M.P. Dubois (INRA, Nouzilly, FranceL The specificity controls of the last two antisera were described previously (11,12L Rat pituitary hormones used as antigens, rLH32V02, rFSH32V02, rPRL32V03, were provided through the National Hormone and Pituitary Program. Antisera were used at following dilutions for immunocytochemistry: Anti-oLHfl, 1:5000 to 1:10 000; anti-pLH/L 1:2000 to 1:5000; anti-oFSHfl, 1:1000 to 1:2000; anti-rPRL, 1:3000 to 1:5000; anti-pTSHfL 1:2000 to 1:5000; antiACTH~.2,, 1:2000 to 1:5000; anti-eGH, 1:5000 to 1:10 000. Immunocytochemical controls. To establish specificity of the staining, one of the major components in the mixture was omitted during the staining procedure. This consisted of using PBS buffer with 1% BSA or 1% normal rabbit serum (Biosys, Compiegne, France} instead of an antiserum, or using normal sheep serum (Institut Pasteur, Paris, France} instead of AIgG-Per. To test specificity of the staining, antisera were incubated with increasing concentrations of various hormones or hormone subunits for 48 h at 4 ~ C. The resulting solutions were then used for staining dispersed pituitary cells. Cell type quantitation. Once stained, the slides were quantitatively analyzed for presence of various cell types. This was done by counting positively stained cells by means of a grid mounted in the ocular lens of the microscope. Usually three slides stained with the same antiserum were chosen. The area of a slide for counting was determined in a random fashion, and a given cell type was counted in an area containing a total population of 800 to 2000 cells. For the cell counting we considered that the cell was positively stained when its cytoplasm was
434
HURBAIN-KOSMATH ET AL.
totally or partly stained. Cells with only peripheral staining were not counted. The black grains observed in
some unstained cells represent silver grains not completely removed in the silver intensification method ~25).
FIG. 1. Morphology of RC-4B/C rat pituitary tumor cell line. a, Light micrograph of 5-/am section of donor ACI/fMai pituitary adenoma. Mitotic figures are present (arrows~. X427. b, Phase microscopy of the RC-4B/C cell line showing the epithelioid nature of the cells. )
