| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Cell Biology, Genentech, Inc. (R.L., A.M., J.P.M.), South San Francisco, California 94080; and The Population Council (D.M.P.), New York, New York 10021
Address all correspondence and requests for reprints to: Dr. Jennie P. Mather, Cell Biology, Genentech, 460 Point San Bruno Boulevard, South San Francisco, California 94080. E-mail: mather.jennie{at}gene.com
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Follicle selection is known to be affected by gonadotropins (for review, see Ref. 2). In cycling animals, small follicles are recruited into the growth pool each cycle by a secondary FSH surge. The development of larger follicles is stimulated by and dependent on FSH. However, the initial growth of granulosa cells in small follicles is independent of FSH and is observed in hypophysectomized animals. Previously, using primary culture, we showed that FSH did not affect the proliferation of granulosa cells from small follicles in vitro and stimulated the proliferation of granulosa cells from large follicles only in the presence of activin (6). This suggested that granulosa cells from different sized follicles were substantially different from each other in their hormone response in vitro. This difference might be the basis for the selective fate commitment.
The study of granulosa cell fate determination in vitro has been difficult due to the lack of homogeneous and stable granulosa cell cultures. Most granulosa cell cultures have been prepared from ovaries that contain follicles of various sizes. In these cultures, a large proportion of the granulosa cells will die spontaneously within a few days. Moreover, granulosa cells in culture can vary widely in their FSH responsiveness with time and culture conditions. To circumvent these problems, we have established an immortalized granulosa cell line in serum-free medium using activin A as a mitogen. The cell line has been grown continuously without FSH and maintains a predifferentiated granulosa phenotype similar to that seen in primary cultures of granulosa cells from prepubertal animals or very immature small follicles. Irreversible cell differentiation can be induced by exposing the cells to FSH. The differentiated cells are postmitotic and will die rapidly upon FSH withdrawal.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell culture
Ovarian follicles were isolated from day 14 Sprague-Dawley rats
as described previously (6). The protocols were approved by the
Genentech Animal Research Facility review committee. The isolated
follicles were plated in 100-mm tissue culture dishes in serum-free
F12-DMEM supplemented with insulin (10 µg/ml), transferrin (5
µg/ml), fibronectin (precoat), and activin A (75 ng/ml). After
24 h of culture, the cells were treated with 1 µg/ml
9,10-dimethyl-1,2-benzanthracene (DMBA) for 1 h and washed three
times with serum-free F12-DMEM (1:1). The cells were cultured in the
same medium for 30 days, with the addition of activin A every 2 days
and a medium change every 6 days. At the end of this period, the cells
were subcultured by removing the cells from the substratum with
collagenase/dispase solution and replating the cells at a 1:2 split.
Thereafter, aprotinin and BSA were added to the medium to prolong the
life of activin.
Cells in secondary cultures have altered attachment properties, in that
they no longer attach to tissue culture-treated polystyrene dishes,
even in the presence of fibronectin. However, they will attach to
laminin or polylysine-coated surfaces (see below). The cells were
routinely carried as suspended aggregates in F12-DMEM supplemented with
8F (insulin, 10 µg/ml; transferrin, 5 µg/ml; 
-tocopherol, 0.1
µg/ml; progesterone, 10 nM; activin A, 75 ng/ml;
aprotinin, 25 µg/ml; BSA, 0.1%; and bovine plasma fibronectin, 5
µg/ml). The addition of BSA improves cell recovery on passaging and
has a "sparing" effect on the activin concentration required, but
is not required for cell growth. The activin concentration was chosen
as optimal for the culture conditions and time of passage used based on
a dose-response curve. Cells did not survive at concentrations of less
than 10 ng/ml activin (data not shown). At 1- to 2-week intervals, the
cells were pooled by centrifugation (1000 rpm, 5 min) and replated at a
1:2 split. The cells have been cultured continuously for more than 30
months.
Cell line characterization
ROG cells were grown in laminin-coated 75-ml tissue culture
flasks in regular 8F medium to allow the cells to attach and spread.
Upon shipment, the flasks were filled with 8F medium, and ovine FSH was
added at 1 ng/ml to promote mitosis. The cells were sent to the Cell
Culture Laboratory of the Children’s Hospital of Michigan (Detroit,
MI) for karyotyping and determination of species of origin by enzyme
analysis and species-specific immunofluorescence. The cell line was
tested periodically for mycoplasma contamination by Hoescht stain,
mycoplasma antibody staining kit, PCR for mycoplasma RNA, and
culturing.
Microscopy
For electron micrographic studies, ROG cells were collected from
routine or FSH-treated suspension cultures by centrifugation (1000 rpm,
5 min). The cells were resuspended in 0.5 ml PBS and transferred into a
1.5-ml microfuge tube. The cells were briefly fixed for 30 sec by
mixing with 1 ml 2.5% phosphate-buffered glutaraldehyde and
immediately pelleted by centrifugation (5000 rpm, 5 min) with a
microfuge. The supernatant was aspirated, and 1 ml 2.5% glutaraldehyde
was added without disturbing the pellets. The fixed cell pellets were
processed for transmission electron microscopy as previously described
(6). Ovaries from 14-day-old rats were excised and immediately fixed
and processed for transmission electron microscopy as a control
tissue.
FSH binding
Granulosa cells were isolated from 14-day-old rat ovaries as
previously described (6). Primary granulosa cells or ROG cells were
cultured in F12/DMEM supplemented with insulin, transferrin, and
activin A at a density of 105 cells/well in laminin-coated
24-well culture plates. Granulosa cells attached and spread on the
substrate within 24 h. The cells were washed twice with binding
buffer (F12/DMEM containing 0.5% wt/vol BSA) and then incubated with
[125I]hFSH in the presence of different concentrations of
cold recombinant hFSH at 37 C for 1 h in the binding buffer. After
incubation, the cells were washed five times with binding buffer and
lysed in 1 M NaOH. The cell lysates were transferred into
test tubes, and the amount of bound radioactivity was counted using a
-counter. The data were analyzed and plotted with aid of the New
Ligand 1.05 program (8).
Analysis of cell apoptosis
ROG cells were plated on poly-D-lysine-coated glass
chamber slides with regular 8F medium with or without ovine FSH (30
ng/ml). After 72 h, all cultures were washed twice and fed with
fresh 8F medium. Then the FSH primed cultures were subdivided into four
groups, each was given 8F alone or 8F with FSH (30 ng/ml), LH (100
ng/ml), or forskolin (5 nM). Cultures were fixed at 0, 1,
2, 3, 4, 5, 6, 7, and 24 h with either phosphate-buffered 2.5%
glutaraldehyde for electron microscopy samples or 1% glutaraldehyde in
3% sucrose solution for TUNEL staining. Changes in cell morphology
were also recorded by time lapse video cinematography.
For DNA fragmentation analysis, ROG cells were grown in 100-mm tissue culture dishes and treated with 30 ng/ml FSH for 72 h. Then, the cells were collected by centrifugation (1000 rpm, 5 min), washed twice with 8F medium, and replated in 8F with or without FSH at a density of 106 cells/dish. At different intervals, cells were harvested and processed for DNA extraction. Extracted DNA was labeled with [32P]didexoy-UTP with a terminal nucleotide transferase labeling kit, fractionated by electrophoresis on agarose gel, and detected by autoradiography. Detailed methods have been described previously (9).
Steroid assay
ROG cells from stock cultures were pooled by centrifugation,
washed twice with F12/DMEM medium, and plated in 35-mm tissue culture
dishes in medium with all factors except progesterone. The cells were
treated with FSH at various concentrations for 72 h. The
conditioned media were collected and assayed for progesterone by RIA.
For assessment of aromatase activity, the cells were treated with
various concentrations of FSH as described above and then incubated in
the presence of 10-7 M androstenedione or
testosterone for another 24 h before collecting the conditioned
medium. The concentrations of progesterone and estradiol in the
conditioned medium were assayed by RIA as previously described (6).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Dispersion of the
cells into single cells will eventually result in cell death. The
karyotype of the cell line confirmed that it was derived from a female
rat with chromosome number in the range of 39–44. Three marker
chromosomes have been identified and determined to be derived from
chromosomes 7, 11, and 12. Two of them, shown in Fig. 1B, were found in
every cell. Only 10% of the cells contained the third marker
chromosome. These marker chromosomes might result from DMBA-induced
recombination.
|
). They had abundant mitochondria and formed numerous
gap junctions. Lipid droplets, a structure frequently seen in more
differentiated granulosa cells, were not observed in these
predifferentiated cells. However, ROG cells showed many dilated
cisternae in the cytoplasm, which may be an adaptation to long term
in vitro culture. Abundant cell processes, usually observed
in cumulus cells, were present in ROG cells.
|
, ROG cells continuously
increased in cell number in 8F medium containing activin. In contrast,
ROG cells plated in the absence of activin failed to grow and
eventually died. Neither transforming growth factor-ß nor inhibin can
substitute for activin A as a survival factor, although activin B does
support cell survival and mitosis (data not shown). Inhibin at
concentrations of up to 1 µg/ml was unable to antagonize the effect
of 30 ng/ml activin (data not shown). The addition of FSH slightly
improved cell survival in the absence of activin, but did not support
long term survival of the cell line. When FSH was given in the presence
of activin, there was a burst of cell growth, but by 72 h after
the addition of FSH, the cells ceased to divide and remained quiescent
thereafter (Fig. 3). Once exposed to FSH, ROG cells were no longer
capable of growing and would not survive through continuous subculture
even when FSH was given repeatedly.
|
) and the ROG cell line (Fig. 4B). Scatchard analysis
of these data showed similar binding affinities in both cultures
(Kd = 2.29 x 10-10 ±
8.10 x 10-11 M-1 and
Kd = 5.77 x 10-10 ± 1.82
x 10-10 M-1 for primary and ROG
cells, respectively). ROG cells had a somewhat higher binding capacity
(2820 ± 780 sites/cell) than the primary culture (1610 ±
450 sites/cell).
|
). ROG cells did not
synthesize detectable (<50 pg/ml) amounts of progesterone in cultures
carried in the absence of FSH. However, de novo synthesis of
progesterone to levels exceeding 500 pg/ml could be induced by 3 days
after the addition of FSH (Fig. 5B). The effective dose range of FSH
for the stimulation of progesterone production was comparable to that
for the stimulation of cell proliferation (Fig. 5B). In contrast, the
production of estradiol by ROG cells was low (
50 pg/ml) even after
FSH stimulation. Measurement of estradiol levels after the addition of
testosterone as a substrate suggests that aromatase activity in ROG
cells may be impaired under these culture conditions (data not
shown).
|
). In the presence of activin, most cells survived as
loosely attached cell aggregates (Fig. 6B). When FSH was given to the
cells without activin, all the cells began to spread and formed a flat
monolayer (Fig. 6C). Treatment with FSH and activin resulted in a
morphology (Fig. 6D) very similar to that seen in hyperstimulated
primary cultures of granulosa cells (also see Fig. 8A). When serum was
added to the cultures, the cell aggregates spread on culture wells
showed a marked change in cell morphology and did not survive
subculture. Cells grown in the presence of serum lost their response to
FSH (data not shown).
|
|
) after FSH withdrawal.
Time lapse recording and scanning electron microscopy showed that the
cells began to retract their cytoplasmic projections immediately after
FSH withdrawal. Cytoplasmic membrane blebbing began after 1 h.
Most cells underwent active membrane blebbing between 2–5 h, retracted
into a smooth apoptotic body, and finally disintegrated into several
pieces. Few cells survived more than 7 h beyond FSH withdrawal.
These changes in cell morphology were more evident when the cells were
viewed using the scanning electron microscope. FSH-treated ROG cells
showed abundant microvilli on their cell surface (Fig. 8A
). The cells rapidly lost their microvilli immediately
after FSH 8 withdraw (Fig. 8B), began to bleb, and then contracted into
apoptotic bodies (Fig. 8, C and D).
|
, the DNA fragmentation was obvious after
4 h of FSH withdrawal. In contrast, the cells that had never been
exposed to FSH and the cells rescued by FSH after FSH withdrawal did
not show DNA fragmentation. TUNEL labeling of the cultures confirmed an
increase in apoptosis 4 h after the withdrawal of FSH (data not
shown). The apoptosis of ROG cells after FSH withdrawal could be
rescued by returning FSH to the cultures immediately (Fig. 7) or by
adding the cAMP activator, forskolin, but not by adding LH (data not
shown). This suggests that these cells do not acquire the ability to
produce cAMP in response to LH after 3 days of FSH treatment.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Serum-free culture is the key to obtaining predifferentiated
granulosa cells
Over the years, several attempts have been made to establish
functional granulosa cell lines. The methods used include spontaneous
immortalization (12), infection with simian virus 40 (13) or oncogenes
(14), and use of antioncogene transgenic animals (15). However, most
cell lines that have been established retain only some of the function
of fully differentiated granulosa cells. Some cell lines retain the
steroidogenic functions but do not show much response to gonadotropins
(16, 17). The basal secretion of steroid hormones and the presence of
LH receptors in these cell lines suggest that these cell lines are
derived from mature granulosa cells.
We approached the problem of establishing granulosa cell lines by combining chemical mutagenesis with selection for the pregranulosa cell phenotype using serum-free hormone-supplemented defined medium. Several laboratories have successfully used serum-free medium supplemented with appropriate growth factors to select for and propagate functional cells of interest (18, 19, 20, 21, 22). Cell lines established in serum-free medium often maintained normal functional responses characteristic of the developmental stage and cell type from which the cell line was derived. For example, Schwann cell lines were able to initiate axon myelination (20). Neural precursor cell lines were capable of neuronal differentiation in vitro and integrated and differentiated into neurons/glial cells in vivo when the cells were implanted into neonatal rat brain (21). To obtain a granulosa cell line, we initiated the primary cultures from immature predifferentiated granulosa cells that would be capable of further replication in vivo. Activin A was used as a mitogen in the culture medium, since granulosa cells from small follicles in primary culture had been shown to require activin for replication (6). Withholding serum from the culture medium allowed ROG cells to continue dividing and maintain the FSH response.
The ROG cell line, derived from P14 rat ovary, maintains many characteristics of predifferentiated granulosa cells. The cell line was free of contamination by other cell types or microorganisms. ROG cells cultured as described in Materials and Methods (cell culture) did not have the phenotype of mature granulosa cells, e.g. steroidogenesis and response to LH. However, as seen with primary cultures of predifferentiated granulosa cells, de novo synthesis of steroid hormones was strongly stimulated when ROG cells were treated with FSH in the presence of activin A.
Early development of granulosa cells depends on activin
Recent evidence suggests that activin may play a major role as a
local regulator in ovarian follicles that both produce and respond to
activins (23) (for reviews, see Refs. 24, 25). The expression of
activin subunit messenger RNA (mRNA) in granulosa cells is limited in
growing follicles, but not in atretic follicles (26). Injection of
FSH/PMSG or diethylstilbestrol markedly increases activin mRNA and
protein levels in hypophysectomized immature female rats, concurrent
with the induction of follicular growth (27). In the human ovary,
activin homodimers were localized in granulosa cells of primary and
secondary follicles as well as in granulosa cells and cumulus cells of
large preovulatory follicles (28).
Activin binding sites (23) and activin receptor mRNAs (11, 29) were found in granulosa cells. We previously reported that activin was a potent mitogen for granulosa cells isolated from day 14 and 21 rat ovaries and suggested that activin was an important factor in the regulation of follicular growth in vivo. The stimulation of cell proliferation by activin was confirmed by other researchers using immature rat granulosa cells (30), sex cord cells of rat embryos (31), and granulosa-luteal cells from human preovulatory follicles (28).
Taken together, this evidence suggested that activin was a mitogen for granulosa cells of all developmental stages in a variety of different species. Moreover, activin was able to induce FSH receptors in immature granulosa cells (10, 11). In the present study, we found that activin supported cell survival and cell proliferation of the ROG cell line and maintained the functional FSH receptors in these cells throughout long term culture in the absence of FSH. It is interesting that the mitotic response to FSH requires the presence of activin, but not that of estradiol, which is not present in these defined cultures. This suggests the possibility that estradiol’s mitogenic effects may also be mediated via the activin autocrine/paracrine loop in the ovary. These data indicate that activin may be an important factor in initiating the growth of primary follicles long before the initial FSH surge.
FSH and follicular selection
Follicular atresia occurs at all stages of follicle growth and
development (32). However, the time during which follicles become most
susceptible to atresia is at the early antral stage (1) when the serum
FSH level is declining after the secondary FSH surge. Increasing serum
FSH levels can rescue follicles from atresia and produce
superovulation. Thus, PMSG was used to establish an experimental
in vivo atresia model. Injection of PMSG initially
stimulated follicle growth, but this was followed by massive atresia as
the PMSG was cleared from the circulation.
Recently, granulosa cell apoptosis has been correlated with, and therefore used as an index for, follicular atresia (3, 5). TUNEL staining has been widely used to identify apoptotic cells. However, TUNEL itself cannot differentiate apoptotic death from late necrosis, where DNA is also degraded (33). In the present study, we employed time lapse video to follow the whole process of granulosa cell death after FSH withdrawal. The membrane blebbing and disintegration of cells into small apoptotic bodies were obvious. When viewed with scanning electron microscopy, the initial change in granulosa cell morphology after FSH withdrawal was impressive. The fast retraction of granulosa cell microvilli immediately after FSH withdrawal provided another indication of granulosa cell death, which occurred much earlier than the membrane blebbing and, of course, the DNA fragmentation. This phenomenon is particularly important because cumulus cells are connected with oocytes by gap junctions on these microvilli. The retraction of cumulus microvilli may isolate oocytes from their surrounding granulosa cells. The significance of this finding to events occurring in vivo awaits further experiments.
A variety of hormones and growth factors have been shown to be able to suppress granulosa cell apoptosis (34, 35, 36, 37). Among them, activin was shown to suppress granulosa cell apoptosis in early antral follicles, and FSH was shown to be a major survival factor for granulosa cells (37). The data presented in this report are in agreement with these observations. As we used a homogeneous population of granulosa cells, we have been able to go one step further and to show the shift of activin-dependent early stage predifferentiated granulosa cells to the FSH-dependent late stage mature granulosa cells. This shift was induced by FSH itself.
In summary, we have developed a novel cell line that has the phenotype of immature or predifferentiated granulosa cells. The ROG cells require activin for survival and growth. These cells maintain normal levels of FSH receptor and are FSH responsive over several years in vitro. However, exposure to FSH rapidly induces the differentiation of the ROG cells to a postmitotic, highly steriodogenic cell phenotype similar to that of mature granulosa cells from a dominant follicle. These cells are now dependent on continuous exposure to FSH for survival and undergo rapid apoptotic cell death on FSH withdrawal. These data suggest that activin and FSH interact to regulate follicular selection and atresia in vivo.
Received October 25, 1996.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-subunit promoter/simian virus T-antigen fusion
gene: characterization of ovarian tumors and establishment of
gonadotropin-responsive granulosa cell lines. Mol Endocrinol 9:616–627This article has been cited by other articles:
 |
|
 |
|
  D. Loo, N. Pryer, P. Young, T. Liang, S. Coberly, K. L. King, K. Kang, P. Roberts, M. Tsao, X. Xu, et al. The glycotope-specific RAV12 monoclonal antibody induces oncosis in vitro and has antitumor activity against gastrointestinal adenocarcinoma tumor xenografts in vivo Mol. Cancer Ther., March 1, 2007; 6(3): 856 - 865. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Ko, N. A. Grieshaber, I. Ji, and T. H. Ji Follicle-Stimulating Hormone Suppresses Cytosolic 3,5,3'-Triiodothyronine-Binding Protein Messenger Ribonucleic Acid Expression in Rat Granulosa Cells Endocrinology, June 1, 2003; 144(6): 2360 - 2367. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. A. Grieshaber, C. Ko, S. S. Grieshaber, I. Ji, and T. H. Ji Follicle-Stimulating Hormone-Responsive Cytoskeletal Genes in Rat Granulosa Cells: Class I {beta}-Tubulin, Tropomyosin-4, and Kinesin Heavy Chain Endocrinology, January 1, 2003; 144(1): 29 - 39. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Nishi, T. Yanase, Y.-M. Mu, K. Oba, I. Ichino, M. Saito, M. Nomura, C. Mukasa, T. Okabe, K. Goto, et al. Establishment and Characterization of a Steroidogenic Human Granulosa-Like Tumor Cell Line, KGN, That Expresses Functional Follicle-Stimulating Hormone Receptor Endocrinology, January 1, 2001; 142(1): 437 - 445. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. A. Grieshaber, S. Boitano, I. Ji, J. P. Mather, and T. H. Ji Differentiation of Granulosa Cell Line: Follicle-Stimulating Hormone Induces Formation of Lamellipodia and Filopodia via the Adenylyl Cyclase/Cyclic Adenosine Monophosphate Signal Endocrinology, September 1, 2000; 141(9): 3461 - 3470. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P. Rose, R. E. Gaines Das, and A. H. Balen Definition and Measurement of Follicle Stimulating Hormone Endocr. Rev., February 1, 2000; 21(1): 5 - 22. [Abstract] [Full Text]
|
|
|
|
 |
|
 H. Zhang, M. Vollmer, M. De Geyter, Y. Litzistorf, A. Ladewig, M. Durrenberger, R. Guggenheim, P. Miny, W. Holzgreve, and C. De Geyter Characterization of an immortalized human granulosa cell line (COV434) Mol. Hum. Reprod., February 1, 2000; 6(2): 146 - 153. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Minegishi, T. Hirakawa, H. Kishi, K. Abe, Y. Abe, T. Mizutani, and K. Miyamoto A Role of Insulin-Like Growth Factor I for Follicle-Stimulating Hormone Receptor Expression in Rat Granulosa Cells Biol Reprod, February 1, 2000; 62(2): 325 - 333. [Abstract] [Full Text]
|
|
|
|
|
|
J.-P. Stephan, P. E. Roberts, L. Bald, J. Lee, Q. Gu, B. Devaux, and J. P. Mather Selective Cloning of Cell Surface Proteins Involved in Organ Development: Epithelial Glycoprotein Is Involved in Normal Epithelial Differentiation Endocrinology, December 1, 1999; 140(12): 5841 - 5854. [Abstract] [Full Text]
|
|
|
|
|
|
N. Walther, A. Einspanier, and M. Jansen Expression of Granulosa Cell-Specific Genes and Induction of Apoptosis in Conditionally Immortalized Granulosa Cell Lines Established from H-2Kb-tsA58 Transgenic Mice Biol Reprod, May 1, 1999; 60(5): 1078 - 1086. [Abstract] [Full Text]
|
|
|
|
 |
|
 R. L. Robker and J. S. Richards Hormone-Induced Proliferation and Differentiation of Granulosa Cells: A Coordinated Balance of the Cell Cycle Regulators Cyclin D2 and p27Kip1 Mol. Endocrinol., July 1, 1998; 12(7): 924 - 940. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
