JP2002510055A - Assay for growth differentiation factor 9 - Google Patents

Abstract

(57) Abstract The present invention relates to a method for identifying a drug that alters (inhibits or enhances) the activity of GDF-9. The method includes a step of combining a GDF-9 receptor and a cell having a gene whose gene expression is regulated by binding of GDF-9 to the receptor, GDF-9; and a drug to be evaluated. The prepared combination is maintained under conditions suitable for binding GDF-9 to the receptor on the cell. The extent to which GDF-9 binding to the receptor on the cell occurs is then measured. In that case, an agent whose GDF-9 binding to a lower or higher degree of receptor alters GDF-9 activity in the presence or absence of the agent to be evaluated than in its absence.

Description

DETAILED DESCRIPTION OF THE INVENTION

Related Application The present application filed with Provisional Application No. 60 / 080,385, filed on April 1, 1998.
Claim priority to the issue. The entire teachings are incorporated herein by reference.

GOVERNMENT ASSISTANCE The present invention relates to NIH Grant (NICHD) RO1 HD33438, GM-0
All or part was assisted by 7330 and GM-08307. The government has certain rights in the invention.

BACKGROUND OF THE INVENTION [0003] Oocytes play a vital role in regulating follicle formation in the ovaries of mammals. In particular, oocytes act on granulosa cells, regulate follicle formation in the perinatal period,
It has been shown to stimulate granulosa cell proliferation, regulate granulosa cell gene expression, and affect steroid synthesis [(Eppig, JJ, Dev. Biol., 5: 51-59 (199
4)]. Granulosa cells in preovulatory follicles can be divided into two populations with respect to their proximity to oocytes; cumulus granulosa cells (cumul)
us granulosa cell) closely surrounds the oocyte, whereas mural granulosa cells are located around the periphery of the follicle separated from the oocyte by the follicular cavity. Cumulus cells (cu
mulus cells secrete a hyaluronic acid-rich matrix during cumulus expansion and are excreted with oocytes during ovulation. This expanded matrix is an important factor for reproductive integrity. Because the matrix binds oocytes and both cumulus cells, follicular extrusion and ovarian fimbrial supplementation (ovi).
because it promotes D. fimbria capture and allows sperm penetration and fertilization [Salustri, A., et al., Zygote, 4: 313-315 (1996)]. On the other hand, mural granulosa cells synthesize a protease important for follicular rupture during ovulation and stay in the ovary after the cumulus cell-oocyte complex is released, eventually undergoing terminal differentiation to the corpus luteum To form These differences in location and function in the granulosa cell population suggest that a gradient of factors secreted by oocytes regulates gene expression and consequent cell differentiation. In vitro studies show that growth factors secreted by oocytes regulate granulosa cell synthesis of hyaluronic acid, urokinase plasminogen activator (uPA) and LH receptors and steroid synthesis and luteinization [ Salustri, A., et al., Zygote 4:31
3-315 (1996); Vanderhyden, BC, et al., Endocrinology, 133: 423-426 (1993); Epipi
g, JJ, et al., Biol.Reprod., 56: 976-984 (1997); Eppig, JJ, et al., Human Repr.
od., 12: 127-132 (1997); Nekola, MV and Nalbandov, AV, Biol. Reprod., 4: 154-
160 (1971); El-Fouly, MA, et al., Endocrinology, 87: 288-293 (1970)]. However, little is known about the properties of oocyte-derived factors that regulate these somatic functions.

[0004] A mouse growth differentiation factor (growth differentiation fa)
ctor) 9 (mGDF-9) is expressed in the ovaries, especially in oocytes (PCT Publication WO 94/15966). GDF-9 knockout mice exhibited female infertility due to an early block of folliculogenesis [Dong et al.
al., Nature, 383: 531-535 (1996)]. However, GDF-9 receptor and cell types expressing GDF-9 receptor have not been identified. Finding a cell type that expresses the GDF-9 receptor will help define the role of GDF-9 in fertility. It will also provide an assay system for identifying agents that target GDF-9 activity.

SUMMARY OF THE INVENTION [0005] Described herein is an in vitro assay for growth differentiation factor 9 (GDF-9) based on the discovery that GDF-9 binds to granulosa cells found in the ovaries of mammals. Also, as described herein, the expression of specific proteins is enhanced and / or inhibited upon binding of GDF-9 to receptors on granulosa cells.

[0006] The present invention relates to methods for identifying agents that alter (modulate) the activity of GDF-9.
As used herein, the term "alter" refers to a partial and / or complete inhibition of activity (decrease in activity) or enhancement of activity (increase in activity). The method includes the use of a GDF-9 receptor and a gene (one or more) whose expression is regulated by the binding of GDF-9 to the receptor [eg, hyaluronan synthase, a steroidogenic early regulator protein (steroidogenic ac).
ute regulatory protein (StAR), luteinizing hormone (LH) receptor, cyclooxygenase 2 (COX-2), urokinase plasminogen activator (uPA), kit ligand, activin /
A cell having inhibin βB and follistatin] (eg, granulosa cell); GDF-9; and a step of combining the drug to be evaluated (test sample). Combinations prepared in test samples are maintained under conditions suitable for binding GDF-9 to receptors on the cells. The extent to which expression of the gene occurs is then measured. At that time, a change in gene expression upon binding of GDF-9 to the receptor in the presence of the drug to be evaluated indicates that the drug alters GDF-9 activity. The method may further include a control step. In doing so, the degree of binding that occurs in the presence of the drug is compared to the degree of binding that occurs in the absence of the drug. For example, the degree of gene expression occurring in a test sample can be determined by comparing a control sample (e.g., a cell having a receptor for GDF-9 and a gene whose expression is regulated by binding of GDF-9 to the receptor) to a combination comprising GDF-9. ) Is compared with the degree of gene expression that occurs in

[0007] The invention also relates to a method of identifying an agent that is an inhibitor of GDF-9 activity. In one embodiment, granulosa cells; GDF-9; and the agent to be evaluated are combined.
The prepared combination is maintained under conditions suitable for binding of GDF-9 to the receptor on granulosa cells; and GDF to the receptor produced upon binding of GDF-9 to the receptor on granulosa cells The extent of expression of the gene regulated by -9 binding is determined. Inhibition of gene expression upon binding of GDF-9 to the receptor in the presence of the drug to be evaluated indicates that the drug inhibits GDF-9 activity.
In another embodiment, the granulosa cells; GDF-9; and the agent to be evaluated are combined. The prepared combination is maintained under conditions suitable for binding of GDF-9 to the receptor on granulosa cells; and GDF to the receptor produced upon binding of GDF-9 to the receptor on granulosa cells -9, a gene regulated by the binding of -9
The degree of expression of A is measured. G to the receptor in the presence of the drug to be evaluated
Increased expression of the uPA gene upon binding of DF-9 indicates that the agent inhibits GDF-9 activity.

[0008] The present invention also relates to a method for identifying an agent that is an enhancer of GDF-9 activity. In one embodiment, granulosa cells; GDF-9; and the agent to be evaluated are combined.
The prepared combination is maintained under conditions suitable for the binding of GDF-9 to the receptor on the granulosa cells, and the GDF-9 on the receptor generated upon binding of GDF-9 to the receptor on the granulosa cells The degree of expression of the gene regulated by the binding of is measured. Enhanced gene expression upon binding of GDF-9 to its receptor in the presence of the drug to be evaluated indicates a drug that is an enhancer of GDF-9 activity. In another embodiment, the granulosa cells; GDF-9; and the agent to be evaluated are combined. The prepared combination is maintained under conditions suitable for binding of GDF-9 to the receptor on the granulosa cells; and The extent of expression of the uPA gene regulated by -9 binding is determined. A decrease in the expression of the uPA gene upon binding of GDF-9 to the receptor in the presence of the drug to be evaluated indicates that the drug enhances GDF-9 activity.

[0009] The methods of the invention can also be used to identify agents that inhibit fertility in mammals (eg, humans). On the other hand, the methods of the present invention can be used to identify agents that enhance fertility in mammals.

[0010] The present invention also relates to a method for identifying a drug that is an agonist of GDF-9. In this method, cells having a GDF-9 receptor and gene, wherein the expression of the gene is regulated by binding of GDF-9 to the receptor; GDF-9; and the agent to be evaluated. GDF-
9 are maintained under conditions suitable for binding, and the extent to which expression of the gene regulated by binding of GDF-9 to the receptor occurs is determined. At that time, expression of the gene in the presence of the drug to be evaluated indicates that the drug is an agonist of GDF-9.

Furthermore, the present invention relates to a method for identifying a drug that is an agonist or antagonist of GDF-9. In this method, a cell having a receptor for GDF-9 and a gene whose expression is regulated by binding of GDF-9 to the receptor; GDF-9
And the drug to be evaluated. The prepared combination is maintained under conditions suitable for binding the drug to the receptor on the cell. The extent to which binding of the drug to the receptor on the cell occurs is measured. At that time, the binding of the drug to the GDF-9 receptor is determined by the
Indicates an agonist or antagonist of F-9.

The extent to which GDF-9 binds to a receptor on a cell can be measured by various methods. For example, the extent to which binding occurs is regulated by the binding of GDF-9 to the receptor.
Nucleic acids; proteins and peptides encoded by genes). In one embodiment, the gene encodes a protein involved in the synthesis of hyaluronic acid (eg, hyaluronan synthase) and determines the extent to which binding of GDF-9 to a receptor on granulosa cells occurs, eg, by the product of the gene (eg, It is determined by measuring the production of RNA encoding hyaluronan synthase). In that case, an increase in the production of the gene product indicates that the drug is an enhancer of GDF-9 activity, and a decrease in the production of the gene product indicates that the drug is an inhibitor of GDF-9 activity. In another embodiment, the gene encodes a protein involved in the synthesis of progesterone (eg, StAR) and determines the extent to which binding of GDF-9 to a receptor on granulosa cells occurs, eg, by the product of the gene (eg,
It is measured by measuring the production of RNA encoding StAR). that time,
Increased production of the gene product indicates that the drug is an enhancer of GDF-9 activity, and decreased production of the gene product indicates that the drug is an inhibitor of GDF-9 activity. In yet another embodiment, the gene encodes a protein involved in plasmin production (eg, uPA) and is regulated by binding of GDF-9 to the receptor that occurs upon binding of GDF-9 to the receptor. The degree of expression of the gene is measured by measuring the production of the product of the gene (eg, RNA encoding uPA). At that time, a decrease in the expression of the gene product indicates that the drug is an enhancer of GDF-9 activity, and a decrease in the expression of the gene product indicates that the drug is a GDF-
9 indicates an inhibitor of 9 activities. The extent to which GDF-9 binding to the receptor on the cell occurs also depends on the product or function resulting from the activity of the protein encoded by the gene whose expression is regulated by the binding of GDF-9 to the receptor on the cell (eg, , Hyaluronic acid, progesterone and / or plasmin).

The finding that granulosa cells respond to GDF-9, and that this response can be measured by one or more inducible genes, allows for an in vitro bioassay of GDF-9. Such assays can be used to diagnose fertility problems in mammals, and for example, inhibitors of GDF-9, which can be used to diagnose and / or treat fertility problems in mammals.
It can be used to identify enhancers, antagonists and analogs.

DETAILED DESCRIPTION OF THE INVENTION The present invention is directed to methods for identifying agents that alter (modulate) the activity of growth differentiation factor 9 (GDF-9). In particular, the methods of the present invention are assays for assessing the ability of an agent to inhibit or enhance the activity of GDF-9, a factor required during early ovarian follicle formation. In a method of the invention for identifying an agent that alters (inhibits or enhances) GDF-9 activity, a test sample is prepared by combining the following reagents: a receptor for GDF-9 and binding of GDF-9 to the receptor. A cell having a gene whose expression is regulated by GDF-9; and a drug to be evaluated.

Cells having a GDF-9 receptor and a gene whose expression is regulated by the binding of GDF-9 to the receptor on the cell include, for example, granulosa cells found in the follicles of mammalian ovaries (eg, , Mural granulosa cells, cumulus cells). Cells with receptors for GDF-9 at all stages of follicle formation (eg,
Ovarian follicles of mammals, including oocytes surrounded by granulosa cells), can also be used in the methods of the invention. For example, Primitive, Primary, Secondary, Tertiary (
Ovarian follicles (before ovulation) or superovulated can be used in the methods of the invention. In this embodiment, the ovarian follicle will generally comprise oocytes and granulosa cells that express GDF-9. Therefore, GDF-9 in the test sample
Is optional in this embodiment. The degree to which the drug to be evaluated inhibits GDF-9 activity or further enhances the activity of GDF-9 present in the ovarian follicles is evaluated. Granulosa cells and / or ovarian follicles can be obtained from any mammalian source as described herein using techniques known in the art. For example, granulosa cells and / or ovarian follicles can be obtained from primate animals (eg, monkeys, humans), cows, pigs, cats, dogs, and murine sources. In addition, granulosa cells and / or ovarian follicles can be obtained from mammals at any stage of follicle formation, or from mammals in a state of superovulation. For example, a condition of superovulation in a mammal can be obtained by administering a gonadotropin (eg, pregnant horse serum gonadotropin) to the mammal, which will provide granulosa cells and / or ovarian follicles. . Other examples of cells suitable for use in the methods of the invention include the GDF-9 receptor and recombinantly produced genes that express a gene whose expression is regulated by binding of GDF-9 to the receptor on the cell. (E.g., recombinantly produced Chinese hamster ovary (CHO) cells that express the GDF-9 receptor and hyaluronan synthase upon binding of GDF-9 to the receptor).

As described herein, the methods of the present invention utilize cells having a receptor for GDF-9 and a gene whose expression is regulated by binding of GDF-9 to the receptor on the cell. Genes (one or more) for use in the present invention include, for example, hyaluronan synthase, steroidogenic early regulatory protein, luteinizing hormone receptor, cyclooxygenase 2, urokinase plasminogen activator, activin / inhibin and follistatin. And a gene encoding

As used herein, “GDF-9” refers to a GDF-9 protein, its individual subunits, multimers of the individual subunits, functional fragments or portions of GDF-9, and GDF-9. Nine functional equivalents and / or analogs. As defined herein, a functional fragment of GDF-9 is
For example, one or more proteins involved in the synthesis of hyaluronic acid (eg, hyaluronan synthase), one or more proteins involved in the synthesis of progesterone (eg, StAR) and / or proteins involved in the synthesis of plasmin ( For example, a fragment that regulates the gene encoding uPA). Also, as defined herein, a functional equivalent or fragment of “GDF-9” includes the resulting GDF-9 product as a GDF-9 as described herein.
Includes GDF-9 proteins that have been modified to have activity similar to DF-9 (eg, regulate genes encoding proteins involved in hyaluronic acid synthesis).

[0018] GDF-9 and compositions of the present invention suitable for use in the present methods can be obtained from a variety of sources, or can be synthesized using techniques known in the art. For example, GDF-9, as occurs in ovarian follicles, can be used in the methods of the invention. In addition, GDF-9 is derived from natural sources (eg, primate animals, cattle,
(Porcine, cat, dog and murine sources), can be purified (isolated, substantially pure), can be produced by chemical synthesis, can be produced by recombinant DNA technology, or can be obtained from commercial sources be able to.

As described herein, the combination (test sample) prepared in the method of the invention is maintained under conditions suitable for binding GDF-9 to a receptor on the cell.
A suitable temperature range for performing the methods described herein is from about 0 ° C to about 45 ° C. In one aspect, the method can be performed at about 37 ° C, and in another aspect, the method can be performed at about 25 ° C. The preferred pH range in which the method can be performed is from about pH 5 to about pH 8, and particularly from about pH 7 to about pH 7.4. In one embodiment, the method can be performed at about pH 7.4. Further, the method can be performed in one day or over several days. For example, the method uses about 1
From about 1 hour to about 48 hours, or from about 1 hour to about 24 hours. In one embodiment, the method can be performed for about 3 hours to about 24 hours.

In the method of the present invention, the extent to which binding of GDF-9 to a receptor on a cell occurs can be measured directly or indirectly. For example, the extent to which binding of GDF-9 to a receptor on a cell occurs to the extent that a particular gene product (e.g., , DNA, RNA (mRNA), peptide, protein). Examples of such specific gene products include:
Including but not limited to hyaluronan synthase, steroidogenic early regulatory protein, luteinizing hormone receptor, cyclooxygenase 2, urokinase plasminogen activator, kit ligand, activin / inhibin βB and follistatin.

On the other hand, the product or function resulting from the activity of the gene product (eg, downstream) can also be measured. In one embodiment, the gene is involved in the synthesis of hyaluronic acid (eg, a gene encoding hyaluronan synthase), and products resulting from the activity of the gene, such as hyaluronic acid, can be measured. In another embodiment, the gene is involved in the synthesis of progesterone (eg, the gene encoding StAR) and produces products resulting from the activity of the gene, such as progesterone, phosphorylated StAR and / or other steroids in the progesterone synthesis pathway. Can be measured. In further embodiments, the gene is involved in plasmin synthesis (eg, uPA), and the product resulting from the activity of the gene, such as plasmin, and / or the function of the gene, such as cleavage of plasminogen, can be measured. In another embodiment, the gene is a prostaglandin (eg, COX-2)
, Which are involved in the production of E. coli and produce from the activity of genes such as prostaglandins. The gene may also be involved in the production of activin B or inhibin B, and products resulting from the activity of the gene can be measured, such as activin βB: activin and / or βB inhibin α: activin βB. In another embodiment, the gene is involved in the production of LH receptor and / or a product resulting from the activity of the gene, such as a product produced as a result of LH receptor binding (eg, cyclic AMP), and / or an LH receptor. Such as binding of a ligand to
The function resulting from the activity of the gene can be measured. Furthermore, in particular, GDF-
In a method where nine agonists or antagonists would be identified, the extent to which binding of GDF-9 to the receptor on the cell in the presence of the drug being evaluated occurs, using techniques known in the art (eg, radioreceptors). Assay)
Can be measured directly by measuring the amount of GDF-9 bound to the receptor on the cell.

The methods that can be used to determine the extent to which binding of GDF-9 to a receptor on a cell occurs depend on the parameters that will be measured. Nucleic acids associated with genes regulated by the binding of GDF-9 to receptors on cells (D
NA, RNA) can be determined, for example, by visualizing the polymerase chain reaction (PCR) product, radiolabeling methods, photographic detection of radiolabeled PCR products, Southern blots of PCR products, RNase protection assays and / or Northern blots. It can be measured using: On the other hand, the presence of proteins related to genes regulated by the binding of GDF-9 to receptors on cells was determined using techniques such as high pressure liquid chromatography (HPLC), immunohistochemistry, Western blot analysis and immunoprecipitation. Can be measured. In addition, G to receptors on cells
Downstream products (eg, hyaluronic acid, progesterone, plasmin) produced in response to the activity of the gene regulated by DF-9 binding can be measured, for example, using HPLC and / or radioimmunoassay.

[0023] The method of the invention may further include the use of a control sample. For example, the extent of GDF-9 binding to receptors on cells in a test sample can be determined by comparing the control sample (ie, including the same combination of reagents as the test sample except for the drug being evaluated,
(A sample that has been processed in the same manner as the test sample) and the extent of GDF-9 binding to the receptor on cells that occurs in the test sample.

The methods described herein and the agents identified by the methods can be used in various ways. For example, the method includes screening agents for identifying agents that are inhibitors, enhancers, antagonists, agonists, and / or analogs of GDF-9 and agents that can be used to diagnose and / or treat reproductive problems. Or as a diagnostic assay to detect reproductive problems such as infertility (eg, an assay of follicular fluid to analyze GDF-9 activity). The agent identified by the method of the present invention is GD
It can be used in any disease or condition where F-9 activity is abnormal. For example, the agents identified in the methods of the invention can be used to inhibit fertility (eg, contraception, contraceptives) or to enhance fertility (eg, increasing successful fertilization in vitro).

As described herein, in early follicle culture experiments using chopped, CHO cell medium conditioned with mature mouse GDF-9 protein (CHO 19A12), the follicles appear granular and It was noted that the cells converged and separated from each other and appeared to form a mucosal secretory cellization matrix after 2 days of culture. CHO control medium or medium containing only bone morphogenetic protein 15 (BMP-15) (CHO 21A5) did not have this mucosal secretion cellizing effect on follicles. Derived from cells expressing both GDF-9 and BMP-15,
In the medium in which -9 levels were estimated to be 5-10 fold lower than in CHO 19A12 (CHO 27A11), a mucosal secretory event occurred after 5 days of culture.

Granulosa cells, which are in direct contact with oocytes in the preovulatory follicle, known as cumulus cells, produce a proteoglycan matrix containing hyaluronic acid both in vivo and in vitro, in a process called cumulus expansion. . This expansion matrix binds the oocytes and both cumulus cells, promotes follicular emptying and fallopian tube trapping, and allows sperm penetration and fertilization (Salustri et al., Zygote, 4: 313-315).
(1996)]. Cumulus expansion in vitro has been observed to be dependent on follicle-stimulating hormone and oocyte-derived factors (Buccione et al., Dev. Biol., 138: 16-25 (19
90)]. Cumulus cells peeled from oocytes do not expand, take the form of adherent fibroblasts, and produce very small amounts of hyaluronic acid. When oocytes are added back to culture, or in oocyte conditioned medium (OCM) (~ 1 oocytes / μl
When growing cumulus cells in culture medium, they produce 5 to 10-fold higher levels of hyaluronic acid. Treatment with oocyte conditioned medium and / or TGF-β induces mural granulosa cells (ie, granulosa cells of preovulatory follicles that have not been contacted with oocytes) and is treated with hyaluronan in vitro. It has also been shown that acids can be made [Salustri et al., J. Biol. Chem., 265: 19517-19523 (1990)]. OCM
The effects of both and TGF-β were additive. Anti-TGF-β antibody is TGF-
The effects of β could be blocked, but the effects of oocyte conditioned media could not be blocked. This indicates that TGF-β itself is not an oocyte-derived signal that stimulates hyaluronic acid production. In addition, these experiments were repeated using actinomycin D, a transcription blocking agent. TGF-β
影響 50% of the effects of A. were shown to be blocked under these conditions [Tirone e
etal., J. Biol. Chem., 272: 4787-4794 (1997)].

Recently, hyaluronan synthase type 2 (HAS2) has been identified in mice and shown to be able to catalyze hyaluronic acid production in transfected COS cells [Spicer et al., J. Biol. Chem. 271: 23400-23406 (1996)]. In another study, a standard superovulation protocol [pregnant horse serum gonadotropin (PMSG)
-48 hours hCG] 1 hour to 4 hours after human chorionic gonadotropin (hCG) administration
After a time, it was shown that HAS2 was shown to be induced in the cumulus oocyte complex [Fulop et al., Arch. Biochem. Biophysics, 337: 261-266 (1997)]. Due to its enzymatic activity and its expression pattern, HAS2 appears to be responsible for hyaluronic acid production leading to cumulus expansion and may be a gene produced by oocytes or responsive to factors present in the oocyte conditioned medium. unknown.

In addition, studies have shown that a second gene, urokinase plasminogen activator (uPA), is inhibited by oocyte-derived factors [
Canipari et al., Devel. Biol., 167: 371-378 (1995)]. uPA is a serine protease that cleaves plasminogen to form plasmin, an active protease, and has been suggested to play a role in pre-ovulation proteolytic breakdown of the follicle wall. uPA is used for granulosa cells and follicular membrane cells (theca).
cell) stimulated by gonadotropin. Cumulus cells in the cumulus oocyte complex do not normally produce uPA. However, cumulus cells that have been stripped from oocytes and stimulated by FSH secrete uPA. uPA secretion
Cultivation of cumulus cells in oocytes or in oocyte conditioned media is greatly reduced.

[0029] Also, as shown in the Examples, recombinants of mGDF-9, an oocyte-specific TGF-β family growth factor, induce HAS2 and / or progesterone expression in granulosa cells, and And / or can inhibit uPA expression. This effect on HAS2 expression is specific for GDF-9 as expected by follicle culture experiments and is induced by other oocyte-specific TGF-β family members (ie, BMP-15 and BMP-6). Not done.

As described in Example 3, mouse GDF-9 protein is expressed in all oocytes and is initiated at the type 3a follicular stage, including follicular cavities. To investigate the biological function of GDF-9 in the late stages of follicle formation, including cumulus expansion,
Mature, glycosylated recombinant mouse GDF-9 was produced using the Chinese Hamster Ovary Cell Expression System. Establish a granulosa cell culture system and use semi-quantitative RT-PCR to regulate GD in the regulation of some key ovarian gene products
The role of F-9 was determined. Recombinant GDF-9 is derived from hyaluronan synthase 2 (H
AS2), cyclooxygenase 2 (COX-2), and steroidogenic early regulatory protein (StAR) mRNA synthesis, but not urokinase plasminogen activator (uPA) and luteinizing hormone receptor (LHR).
) Found to inhibit mRNA synthesis; StAR mR by GDF-9
Consistent with the induction of NA, recombinant GDF-9 increased progesterone synthesis in granulosa cells in the absence of FSH. Since the induction of HAS2 and suppression of the protease uPA in cumulus cells are important events in the production of the hyaluronic acid-rich extracellular matrix produced during cumulus expansion, can GDF-9 mimic this process? I decided. Using an oocyte-excised cumulus cell-oocyte complex, recombinant GDF-9 was shown to induce cumulus expansion in vitro. These studies indicate that GDF-9 can bind to receptors on granulosa cells and regulate the expression of many gene products. Thus, in addition to serving important functions as growth and differentiation factors during early follicle formation, GDF-9 function as a paracrine factor secreted by oocytes, which regulates several key granulosa cell enzymes, Involved in the maintenance of an optimal oocyte microenvironment, a process essential for normal ovulation, fertilization, and female reproduction.

As described in Example 4, the molecular defects resulting from the absence of GDF-9 were analyzed. The main findings were: 1) Several follicular membrane cell layer markers (ie, 17α hydroxylase, luteinizing hormone receptor (LHR)).
, And c-kit, which is a receptor for kit ligand), GDF-9
There is no detectable signal around the defective follicle. This indicates that in the absence of GDF-9, the follicle cannot emit signals that recruit follicular cell precursors to surround the follicle. 2) Primary follicles of GDF-9 deficient mice show up-regulation of kit ligand and inhibin α. This indicates that these two important secreted growth factors, expressed in granulosa cells, are likely to be regulated by GDF-9 directly in a paracrine manner. Also, up-regulation of kit ligand via signaling through c-kit on oocytes may be directly involved in increasing the size of GDF-9 deficient oocytes and the final disappearance of oocytes. It is. 3) After oocyte loss, cells of the GDF-9 deficient follicle remain in steroidogenic clusters that are histologically similar to the small corpus luteum. However, at the molecular level, these cells are positive for both luteal markers (eg, LHR and P-450 side-chain truncation) and non-luteal markers (eg, inhibin α and P-450 aromatase). This indicates that, initially, the presence of the oocyte prevents the expression of the luteinizing marker, whereas the absence of GDF-9 at an earlier time alters the differentiation program of granulosa cells. 4) As shown by staining with either PCNA or Ki-67 and TUNEL labeling, GDF-9 deficient primary 3b type follicular granulosa cells do not proliferate, but do not lead to cell death. This indicates that granulosa cells of type 3b follicles require GDF-9 for continued proliferation and for example to become apoptotic through differentiation events.

The present invention is illustrated by the following examples. Such examples are not intended to be limiting.

[0033]
EXAMPLES Example 1 Granulosa Cell Culture In Vitro Assay for GDF-9 Function Protocol Twenty-one day old CD-1 mice were harvested from Charles River Laborat.
ories or Baylor College medical check-up facilities (Medic
Ine barrier facility colony). Each female was injected with 7.5 IU of pregnant horse serum gonadotropin (PMSG) (0.3 cc of our stock). First, after 48 hours (42 in the last experiment of Example 1)
After ４６46 hours), the mice were sacrificed by cervical dislocation and the ovaries were harvested, removing as much fat and sac tissue as possible. Recovery medium (DMEM / F12 w / 1X GPS and 0.3% BS) in a pre-warmed 30 mm dish
In A), the ovaries were placed. Granulosa cells were released from large follicular follicles by repeated stabbing with a 271/2 gauge needle based on the anatomical field. The aggregate of granulosa cells was dispersed by sucking with a Pasteur pipette and pipetting up and down with the mouth. Oocytes and cumulus oocyte complex (COC) were manually separated from granulosa cells. Once, all obvious oocytes and COC's were removed and the granulosa cells were further transferred to a clean dish with pre-warmed recovery media to remove any remaining oocytes and COC. Transfer the contents of the second dish to two 1.5 ml Eppendorf tubes, 150 minutes for 3 minutes
Rotated at 0 rpm. The remaining cells were washed off the plate using fresh recovery medium, transferred to an Eppendorf and spun again at 1500 rpm.

For each collected ovary, 1 × follicle culture medium (alpha MEM, 1 × insulin / transferrin / selenite, 1 × glutamine / penicillin / streptomycin, 4 ng / ml sheep FSH, +/− 10% fetal bovine serum ) 2
Granulosa cells were resuspended by gentle pipetting from 50 μl to 500 μl and 250 μl of granulosa cells were placed per well of a 24-well plate. For each well, a 2 × volume of 1 × follicle culture medium 250 μL of the treatment under study
l or more. The final culture volume was 500 μl.

Experimental Example 1 Medium: 1) CHO control (concentrated) = 100 μl of concentrated medium derived from CHO A2 (CHO cells without an expression vector) +400 μl of 1 × follicle culture medium (FCM)
. Final dilution = 1: 10. 2) mGDF-9 (enriched) = enriched CHO 19A derived from mGDF-9 expressing cells
12 medium 50 μl + 1 × F. C. M. 450. Final dilution = 1: 20. Estimated final mGDF-9 concentration = 50 ng / ml. 3) mBMP-15 (enriched) = enriched CHO 2 derived from mBMP-15-expressing cells
100 μl of 1A5 medium + 1 × F. C. M. 400 μl. Final mB estimated
MP-15 concentration ≧ 50 ng / ml. 4) mGDF-9 + mBMP-15 (concentrated) = concentrated CHO 19A12 medium 5
0 μl + concentrated CHO 21A5 medium 100 μl + F. C. M. 350 μl. Estimated final mGDF-9 concentration = 50 ng / ml. ^* 1X F. remaining. C. M. 500 μl + cells were spun at 1500 rpm for 3 minutes and 2 × F. C. M. And 250 μl was placed in the two remaining wells. 5) ^* mGDF-9 (no concentration) = 250μ without CHO 19A12 concentration.
2X F.l with l + granulosa cells C. M. 250 μl. Final mG estimated
DF-9 concentration = 300 ng / ml. 6) ^* m dimer (no concentration) = 250 μl without CHO 27A11 concentration (m
CHO cells expressing both GDF-9 and mBMP-15). Estimated final mGDF-9 concentration <150 ng / ml. Estimated final mBMP-15 concentration> 150 ng / ml. All cells were cultured overnight (〜20 hours) at 37 ° C. with 5% CO ₂ .

The medium is aspirated and RNA Stat-60 (Ledo Medical Su)
pply) was added in a single addition to each well and the cells were lysed by pipetting up and down. All cells were lysed with confirmation using a dissecting microscope. The cells were pipetted up and down using a P200 Pipetman to homogenize the cells. 100 μl of chloroform was added to each tube, vortexed and then incubated on ice for 10 minutes. The tube was spun at maximum speed for 20 minutes at 4 ° C. The aqueous phase was carefully collected and the RNA was precipitated with 250 μl of isopropanol at −20 ° C. for 2 hours. After spinning at maximum speed for 30 minutes at 4 ° C., a very small RNA pellet was observed. The pellet was washed with 70% ethanol, air dried, then resuspended in 100 μl of DEPC water and stored at −80 ° C.

GIBCO Superscript II Reverse Transcriptase Kit and Oligo-d
CDNA was synthesized from 5 μl of each RNA sample (１ ／ of total recovery) by reverse transcription (RT) using T primers. The total reaction volume was 20 μl. HAS2 was detected by PCR with the following primers: HAS2-1F (5 ′), which gives the expected product of 430 bp from spliced mRNA
= GCTTGACCCTGCCTCATCTGTGG (SEQ ID NO: 1) and H
AS2-1R (3 ′) = CTGGTTCAGCCATCTCAGATATT (SEQ ID NO: 2). GAPDH was used as a control for RNA integrity and was detected with the following primers: GAPDH-1F (5 ′), giving a 486 bp product.
= CATGTTTGTTGATGGGTGTGAACC (SEQ ID NO: 3) and G
APDH-1R (3 ′) = TGGGAGTTGCTGTTGAAGTCCGCA (
SEQ ID NO: 4). Standard 50 μl P for each primer set under the following amplification conditions
In the CR reaction, 2 μl of each RT reaction was used: 3 min at 94 ° C., 28 × (94 ° C. 1
Min, 60 ° C for 30 seconds, 72 ° C for 1 minute and 72 ° C for 7 minutes). The products were run on a 1.5% agarose gel and visualized.

Results: HAS2 expression was significantly increased (-10-fold) in samples containing mGDF-9 compared to the CHO control and all other growth factors tested.

Experimental Examples 2 and 3: This experimental example was repeated two more times, with controls, mGDF-9 and mBMP-15.
At 5 h and> 20 h of culture in culture, HAS2 induction was seen, and it was consistently seen that HAS2 expression was elevated in mGDF-9 treated mural granulosa cells. Furthermore, this effect of mGDF-9 is not dependent on the presence of serum in the medium or PMSG stimulation of follicular growth in vivo.

Experimental Example 4: mGDF-9 and hGDF in both unconcentrated and enriched conditioned media
-9 was measured by Western blot. MGDF-9 (concentrated) medium is ~
1 μg / ml, mGDF-9 non-concentrated medium contained ０．６0.6 μg / ml. h
GDF-9 conditioned medium contained １1 μg / ml. Final concentration 10 ng / ml, 5
The following experiments were performed using unconcentrated mGDF-9 medium diluted to 0 ng / ml and 100 ng / ml. hGDF-9 conditioned medium, 50 ng / ml,
Tested at 100 ng / ml and 300 ng / ml. The purpose was to define an initial time course for the activity concentration range and effect.

Protocol The following protocol for this experimental example has the following modifications and is identical to that described above.

Since the 1 × follicle culture medium did not contain fetal calf serum, all the effects seen were serum independent.

To increase the size of the precipitated RNA pellet and increase the reliability of the precipitation step, 10-20 ng of yeast tRNA was added to each sample during the homogenization step.

An internal control for RNA integrity and quantification was performed using H with the following primers:
Changed to PRT: HPRT-1F (5 ′) = CCTGGTTAAGCAGTA
CAGCC (SEQ ID NO: 5) and HPRT-2R (3 ') = TACTAGGC
AGATGGCCACAG (SEQ ID NO: 6).

Urokinase plasminogen activator (uPA) levels were tested in each sample, and HAS2 was tested using the following primers: uPA-
3F (5 ') = GTTCAGACTGTGAGATCACTGG (SEQ ID NO: 7
) And uPA-4R (3 ′) = CAGAGAGGACGGTCAGCATGG
(SEQ ID NO: 8).

To increase the sensitivity of product detection, 0.1 μl of αP ³² -dCTP was added to each PCR reaction. Amplification was performed for 20 cycles, using only 1 μl of the RT reaction as a template. 5 μl of each reaction was run on a 4% polyacrylamide gel, the gel was dried and exposed to Kodak X-OMAT autoradiography film.

Results Up to 3 hours in culture, 100 ng / ml mGDF-9 stimulated HAS2 production and inhibited uPA production, while 100 ng / ml hGDF, compared to CHO controls.
DF-9 is unable to stimulate HAS2 and significantly induces uPA. After 5 hours of culture,
10 ng / ml of mGDF-9 induces HAS2, while both 50 ng / ml and 100 ng / ml induce about 5- to 10-fold higher levels. After 5 hours of culture, 100 ng / ml hGDF-9 induces a detectable HAS2 signal, while 300 ng / ml induces levels approaching those of 50 ng / ml mGDF-9. At 5 hours, 100 ng / ml mGDF-9 inhibits uPA expression, while all concentrations of hGDF-9 appear to have a stimulatory effect. These results indicate that the maximal activity of mGDF-9 to induce HAS2 is 5
Near 0 ng / ml, hGDF-9 also suggests that it can bind the mouse GDF-9 receptor and stimulate HAS2 expression, but may be very weak in this in vitro model system. uPA is inhibited by high concentrations of mGDF-9 (at least about 100 ng / ml) but may be stimulated by low concentrations (at most about 50 ng / ml). These observations indicate that oocyte-derived GDF-9 has the lowest level reached by the mural granulosa cells furthest from the oocyte, defining that follicular wall collapse for ovulation will occur. GDF-9 concentration gradient at the follicular cavity caused by the diffusion of
Suggests that GDF-9 regulation of PA may be dependent.

Conclusions Together, these results indicate that mGDF-9 is an oocyte-derived growth factor that normally produces hyaluronic acid production and inhibition of uPA in cumulus cells. Furthermore, these results indicate that granulosa cells respond to GDF-9 and GDF-
Indicates that the 9 receptor can be expressed. See table.

[0049]

[Table 1]

Example 2 Identification of the Stimulatory Effect of Recombinant mGDF-9 on Progesterone Synthesis by In Vitro Cultured Mural Granulosa Cells Protocol Primary granulosa cells were isolated as described above and 10% fetal calf as indicated. Cultures were performed for 5 or 24 hours at various concentrations of recombinant mGDF-9 in αMEM-based medium containing 2-5 ng / ml FSH with or without serum. The medium was removed after culture and contaminating non-adherent cells were removed by centrifugation. Medium at -20
Frozen at ℃. Progesterone levels in the medium were determined by radioimmunoassay (RIA).
). Briefly, a standard amount of radiolabeled progesterone is
Competes with the progesterone contained in the experimental sample for the a-progesterone antibody binding site contained in the RIA tube. A given standard allows the construction of a standard curve that correlates to the number of counts binding to the progesterone level of the experimental sample.
The progesterone RIA kit is available from Diagnostic Products Co.
0.1 to 4 using a 100 μl volume of sample obtained from
A progesterone level of 0 ng / ml is detected. Higher concentrations can be measured by diluting the experimental sample with a given diluent prior to the assay.

Results Recombinant mGDF-9 stimulated progesterone synthesis in a time- and concentration-dependent manner by primary granulosa cells cultured in vitro (see Figures 1, 2 and 3). At 5 hours, 60 ng / ml of mGDF-9 showed a 2.5-fold increase and 300 ng / ml showed a 4.5-fold increase (FIGURE
2). This stimulation of progesterone was also seen at 24 hours, where a 3.5- to 4-fold stimulation was observed (see Figure 3).

DISCUSSION The rate-limiting step in progesterone synthesis depends on the enzyme complex cytochrome P450 side chain cleavage (
Scc) catalysis of cholesterol on pregnenolone. Non-luteinized cultured rat granulosa cells can be stimulated to increase P450scc mRNA levels and produce progesterone in vitro with the cAMP agonist forskolin (Oonk et al., J. Biol. Chem., 264: 219).
34-21942), indicating that progesterone synthesis at this stage is cAMP-dependent and associated with increased P450scc transcription. The results described herein demonstrate that one other function of GDF-9 (ie, stimulation of progesterone synthesis) and one other downstream target gene of GDF-9 signaling (ie, the cytochrome P450 side chain cleavage gene) ). Furthermore, the measurement of progesterone is based on G
Provides a fast and easy functional assay for GDF-9 function for testing DF-9 analogs and inhibitors.

Example 3 Paracrine Action of Growth Differentiation Factor-9 in Mammalian Ovaries Materials and Methods Immunohistochemistry Ovaries from CD1 (ICR) mice (Charles River Laboratories (
Charles River Laboratories collection from Baylor College of Medicine (Baylor College)
of Medicine) was fixed in 10% neutral buffered formalin, processed,
Embedded in paraffin. After dewaxing 4 μm ovarian sections, a series of stepwise
It was rehydrated with a ethanol solution. 0.1 M phosphate buffered saline (PBS) and 0
. 1x wide-area blocking buffer diluted with 1% BSA (Biogenex (B
iogenex), San Ramon, CA for 30 minutes before ink
Non-specific binding was reduced. Mouse anti-primary antibody
Human GDF-9 monoclonal antibody was purchased by Dr. Neil Wolfman
(Genetics Institute, Cambridge
(Massachusetts, Mass.) And a final concentration of 30-60 ng was added to each section.
/ Μl for 2 hours. Sections were washed with 0.1% BSA in PBS for 5 minutes
After washing twice, biotinylated goat anti-mouse IgG (Biogenex, Sun
(Ramon, CA) for 20 minutes. After the above washing,
Sections were stained with streptavidin conjugated to alkaline phosphatase (Biogen
Nex, San Ramon, CA) for 20 minutes.
Wash twice in 0.1% BSA in BS, add New Fuchsin Alka
Along with the rephosphatase substrate, according to the manufacturer's instructions (Biogenex)
And incubated. After detection of a positive reaction, the sections were counterstained with hematoxylin.
Glycerol.

Production of Recombinant Mouse GDF-9 Full-length mouse GDF-9 cDNA (Incerti, B. et al., Biochim. Biophys.
Acta., 1222: 125-128 (1994)), using the processing gene PACE (Genetics Institute, Cambridge, Mass., Mass.).
The gene was subcloned into an expression vector pHTop containing the gene (donated by Dr. Monique Davies). The GDF-9 expression vector was transfected into CHO cells under standard conditions (Gibco BRL Life Technologies) in lipofectin. Expression of mouse GDF-9 in CHO cells was subsequently driven by a tet-regulated promoter, while the SV40 promoter regulated PACE expression. 10% heat-inactivated dialyzed fetal calf serum, 100 μg
/ Ml G418-sulfate (Gibco BRL Life Technologies) and clones stable and positive in the presence of 0.02 μM methotrexate in α-modified Eagle's medium (αMEM) containing the antibiotics gentamicin, penicillin and streptomycin Was elected. After clonal selection and expansion in 0.02 μM methotrexate, 100 mg / ml heparin (Sigma, St. Louis,
GDF-9 in Opti-MEM low serum collection medium containing Missouri).
The expressing cells were incubated for 24 hours. The medium was collected and GDF-9 protein levels were measured by SDS-polyacrylamide gel electrophoresis (PAGE) followed by immunoblotting (see next section).

Western Blot Analysis A sample of the GDF-9-containing medium was subjected to Biorad Mini-Sell (Biorad Mini-S
ubcell) on a 5% enrichment / 15% separation SDS polyacrylamide gel in an apparatus (Sambrook, J., et al., "In: Molecular Cloning": A Laborator).
y manual, (N. Ford and C. Nolan, eds) Vol. 3, 2nd Ed., 3 vol., Cold Spri
ng Harbor Laboratory Press, Cold Spring Harbor, USA (1989)) and subsequently transferred to a PVDF membrane. The membrane was washed with 0.05% Tween-2.
Blocked overnight in 5% non-fat milk in 1 × Tris buffered saline containing 0. Mouse anti-human GDF-9 monoclonal antibody (above) in blocking solution 1
: Anti-mouse secondary antibody conjugated to horseradish peroxidase at a dilution of 1000 (Southern Biotec Associates)
hnology Associates), Birmingham, AL) in blocking solution 1:
Used at 2500 dilution. Chemiluminescent and autoradiographic films using ECL Western detection reagent (Pierce, Rockford, Ill.) Detected the signal. Densitometer (Molecular Dynamics (Mo
lecular Dynamics) and Imagequant software to quantitate the bands and compare the signal intensity of GDF-9 in the conditioned medium to that of a known concentration of GDF-9 standard measured simultaneously. GDF- in conditioned medium
9 were measured. Several batches of recombinant mouse GDF-9 occurred during the course of this experiment and they all had similar activities based on Western blot quantification (
(I.e., immunoreactivity that correlates with biological effects).

Isolation and Culture of Granulosa Cells
. Inject 5 IU of Gestil (Diosynth, BV, The Netherlands) and after 44-48 hours collect the ovaries and excise without fat and peripheral cells,
0.3 mg / ml L-glutamine, 100 U / ml penicillin, 0.1 mg
Per ml of streptomycin (Gibco BRL, Grand Island, NY) and 25 mM HEPES supplemented with 0.3% BSA (Sigma).

The mural granulosa cells carefully remove the oocytes and cumulus cell-oocyte complex (COC) released by puncturing the follicles of the large follicles (see below), Granulosa cells from the cells were pooled, centrifuged, and in the presence or absence of 20% fetal calf serum (Hyclone Laboratories, Logan, UT) and sheep FSH (National Hormone and・ Pictuary program (National Hormone and Pituitary Program)
NIDDK-o-FSH-2 courtesy of Dr. Parlow
2x granulosa cell culture medium (GCM) in the presence or absence of 0): 0.6 mg
/ Ml L-glutamine, 200 U / ml penicillin, 0.2 mg / ml streptomycin and 2x insulin transport selenite (Gibco BRL) in αMEM (Gibco BRL). GDF-9 containing medium or control conditioned medium was diluted to 2x final concentration in α-MEM. An equal volume of 2 × GDF-9 containing medium or control medium was combined with granulosa cells in 2 × culture medium and cultured at 37 ° C. in a humidified atmosphere containing 5% CO ₂ . After culturing for various periods of time, the non-adherent cells were pelleted with medium and the medium was stored at -20 ° C. The granulosa cells are lysed, and the total RNA is converted to RNA Stat-60 (Read Medical Laboratories (
Leedo Medical Laboratories), Houston, Tex.) According to the manufacturer's protocol.

Semi-quantitative RT-PCR analysis The oligo-dT primed cDNA from each RNA sample was
Synthesized using erscript reverse transcriptase (Gibco BRL) according to the manufacturer's protocol. 1 μl of each RT reaction (1/20 of the total) was used in each 25 μl PCR reaction primed with gene-specific oligonucleotide. Mouse HAS2 mRNA expression spreads to 1.4 kb intron 5'GCTT
GACCCTGCCTCATCTGTGG 3 '(SEQ ID NO: 1) (sense) primer and 5' CTGGTTCA
GCCATCTCAGATATT 3 '(SEQ ID NO: 2) (antisense) primer (Fulop, C.,
et al., Arch. Biochem. Biophys., 337: 261-266 (1997)). 4
The 03 bp PCR product is amplified from RNA and is easily distinguished from amplification of contaminating DNA. Mouse uPA mRNA expression was performed using a 5 'GTTCAGACTGTGAGATCACTGG 3' (SEQ ID NO: 7) (sense) primer and a 5 'CAGAGAGGACGGTCAGCATGG 3' (SEQ ID NO: 8) (antisense) primer spanning two introns of 1.4 kb in length. Detected. The 434 bp PCR product is amplified from RNA. Mouse hypoxanthine phosphoribosyltransferase (HPRT) spreads over three introns of unknown size and is 309 bp of putative mRNA from RNA.
Detection was performed using a 5 'CCTGGTTAAGCAGTACAGCC 3' (SEQ ID NO: 5) (sense) primer and a 5 'TACTAGGCAGATGGCCACAG 3' (SEQ ID NO: 6) (antisense) primer, which give a derived product. Mouse StAR mRNA expression spreads over a number of small introns and gives a 522 bp product from the mRNA 5'TCGCTTGGAGGTGGTGGT
AGAC 3 '(SEQ ID NO: 9) (sense) primer and 5' GCAGGTCAATGTGGTGGACAGT
Detection was performed using a 3 ′ (SEQ ID NO: 10) (antisense) primer. Mouse cholesterol side chain truncated P-450 mRNA expression was 5 'GCCAACATTACCGAGATGC 3'
(SEQ ID NO: 11) Detected using the (sense) primer and 5′CGAACACCCCAGCCAAAGCC 3 ′ (SEQ ID NO: 12) (antisense) primer, and
Gives a 26 bp product. Mouse COX-2 mRNA expression is 5'CTCCTTTTCAACCA
GCAGTTCC 3 ′ (SEQ ID NO: 13) (sense) primer and 5′TCTGCAGCCATTTCCT
TCTCTC 3 '(SEQ ID NO: 14) detected using (antisense) primer
Gives a 77 bp product. Mouse LH receptor mRNA expression spreads over a number of introns and gives a 516 bp product 5′CTTATACATAACCACCATACCAG 3 ′ (SEQ ID NO: 15) (sense) primer and 5′ATCCCCAGCCACTGAGTTCATTC 3 ′ (SEQ ID NO: 16) (antisense) primer Detected using Granulosa cell cDNA
PCR products amplified from were first isolated, subcloned, sequenced and confirmed that they matched the published sequence. In later experiments, [α ³² P] −
dCTP was added to each PCR reaction and the products were separated by electrophoresis on a 4% polyacrylamide gel. The gel was dried and subjected to autoradiography, and the radioactive band was analyzed using Molecular Dynamics Phosphor Imager (Molecular
Quantification on a Dynamics Phosphorimager (Storm 860).

Northern Blot Analysis Total RNA was isolated from granulosa cells and analyzed by fluorimetry using Ribogreen RNA quantification reagent (Molecular Probes, Eugene, OR).
An 85-495 nm excitation filter and a 515-525 nm emission filter were used.
Quantification was performed with a VersaFluor fluorometer (BIORAD, Hercules, CA). 15 μg of total RNA from each sample was electrophoresed on a 1.2% agarose / 7.6% formaldehyde gel and transferred to a Hybond N nylon membrane (Amersham, Arlington Heights, IL). StripEZ probe synthesis kit (Ambion
Probes for HAS2 and uPA were generated from the subcloned PCR product by random priming with [α ³² P] -dATP using Austin, TX). Membranes were hybridized, washed, and subjected to autoradiography as described (Mahmoudi, M. and Lin, VK, Biotec
h., 7: 331-332 (1989)). The probe was removed from the membrane using Strip-EZ removal reagent (Ambion) according to the manufacturer's protocol. The same blot was then reprobed with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a loading control. The signal of each probe was quantified with a Molecular Dynamics Phosphor Imager.

Progesterone Radioimmunoassay Specific proprietary solid-phase radioimmunoassay using a kit from Diagnostic Products Corporation (Los Angeles, Calif.) According to the manufacturer's instructions for the determination of progesterone in the medium by 2 Measured times. The sensitivity of this assay is 0.02 ng / ml, 0.1-40 ng / ml.
ml of calibration standard was used.

Expansion of Oocytectomized Complex The cumulus cell-oocyte complex was recovered as described above. Oocytes were removed from each complex using a microsyringe as described above (Buccion).
ione, R.) et al., Dev. Biol., 138: 16-25 (1990)). Successful oocyte resection was evaluated by removing germinal vesicles along with most of the egg quality. 100ng / ml GDF-
Eggs in 20 μl droplets of granulosa cell culture medium supplemented with 10% fetal calf serum and 5 μg / ml or 100 ng / ml oFSH with or without 1 μg / ml oLH in the presence or absence of 9 The maternal resection complex was incubated together for 18 hours. Pictures were taken with a Nikon inverted microscope.

Results Immunohistochemical detection of GDF-9 in mouse ovaries GD was detected in mouse ovaries using a monoclonal antibody against human GDF-9.
The F-9 protein was specifically detected. GDF-9 immunoreactivity was detected only in oocytes in low magnification immunohistochemically stained ovaries, while oocytes in GDF-9 deficient ovaries did not stain. Primitive (type 2) oocytes were negative consistent with the absence of GDF-9 mRNA expression (McGrath, SA, et al., Mol. Endoc.
r., 9: 131-136 (1995)) and Elvin and Matzuk, unpublished data). Initially low (and variable) levels of GD in oocytes of type 3a follicles (follicles with less than 20 cubic granulosa cells arranged concentrically around the oocyte)
F-9 immunoreactivity was observed, and at higher levels in type 3b and subsequent follicular oocytes. The fully matured oocytes of the multilayered pre-alveolar follicle are consistently more strongly stained for GDF-9, and the GDF in the oocytes of the large and preovulatory follicle cumulus-oocyte complex -9 immunoreactivity was clearly detected. As expected for the secreted peptide, GDF-9 immunoreactivity was excluded from the germinal vesicle (eg, nucleus). Interestingly, asymmetric staining patterns were frequently observed in oocytes, presumably due to the detection of precursor GDF-9 in the oocyte endoplasmic reticulum and Golgi complex [Wasserman, PM] And Albertini, DF, "The mammalian ovum", In: E. Knobil
) And Nail (J. Neill) (ed.) The physiology of reproduction, Raven Press,
New York, pp. 79-122 (1994)]. Thus, GDF-9 mRNA and protein are synthesized by oocytes of all mature follicles, which indicates that
folliculogenesis).

Production of Recombinant Mouse GDF-9 To study the effects of GDF-9 at later stages of follicle formation, recombinant GDF-9
It was necessary to produce Chinese hamster ovary (CHO) cells were stably transfected using an expression vector containing both full-length mouse GDF-9 cDNA and cDNA for PACE, a prepropeptide sequence cleavage enzyme. Western blot analysis of medium derived from CHO cells containing the mouse GDF-9 expression vector was analyzed under denaturing conditions and was not present in the mock conditioned medium.
Two distinct bands with molecular weights of about 21 kD and 60 kD were shown. Four bases RRR
Proteolytic processing of GDF-9 at the -R site would be expected to result in a 135 amino acid carboxyl-terminal mature peptide with the expected molecular weight of about 15.6 kD (McPherron, AC) and Lee ( Lee, S.-J
.), J. Biol. Chem., 268: 3444-3449 (1993)). Mature GDF-9 sequence contains N-linked glycosylation sites of single strand ^{(Asn 325 -Leu 326 -Ser 327)} . The predominant 21 kD form would correspond to the cleaved mature monomeric form of mouse GDF-9 with one N-linked oligosaccharide. This band migrated at the same position as recombinant human GDF-9 synthesized in CHO cells. The 60 kD band corresponds to the glycosylated, unprocessed (prohormone) form (441 amino acids), which is also secreted into the medium. To confirm that the increase in molecular weight of these two types was due to N-linked glycosylation, GDF-9 containing media was treated with N-glycanase to remove N-linked oligosaccharides. This process reduces the size of the 21 kD band to 1
It reduced to 6 kd, the same molecular weight as the mature GDF-9 peptide produced by bacteria, and reduced the 60 kD band to 50 kD. 21 kD glycosylated GDF
This strategy of producing recombinant glycosylated GDF-9 in mammalian cells in the presence of PACE and under serum-free media conditions is efficient because type-9 is always the most abundant type.

Hyaluronan synthase 2 (HAS2) and uPAmR with recombinant GDF-9
Regulation of NA Synthesis Oocytes stimulate growth of hyaluronic acid, inhibit urokinase plasminogen activator (uPA), and secrete growth factors known to cause cumulus expansion in vitro (sarcitri (Salustri, A.) et al., Zygote, 4: 313-315 (
1996)). Hyaluronan synthase 2 (HAS2) is expressed in the mouse cumulus cell-oocyte complex after a gonadotropin surge and just prior to efficient cumulus expansion (Flop, C. et al., Arch. Bioch. Bioph. ., 337: 261266 (1997)), which is included as the major hyaluronan synthase involved in cumulus expansion (Spiser (Spi
cer, AP) et al., J. Biol. Chem., 271-23400-23406 (1996)). To determine if GDF-9 is an oocyte secreted factor that increases hyaluronic acid matrix synthesis and induces cumulus expansion through reducing hyaluronic acid matrix degradation, recombinant mouse GDF- The expression levels of HAS2 and uPA were examined in freshly isolated granulosa cells treated with 9. RT-PCR incorporation of radiolabeled nucleotides into specific products was used to monitor HAS2 and uPA expression levels in control and GDF-9 treated granulosa cell cultures. First, a linear range of product amplification was established for each oligonucleotide pair of the three genes (ie, HAS2, uPA and HPRT). Radiolabel [α- ³² P] in PCR reaction
Using dCTP, the same samples were amplified in 16-22 cycles and quantified using photodensitometric analysis of autoradiographic films. The linear increase of the amplification product was 16-20 cycles for HPRT, 16-20 cycles for HAS2, and 18-22 cycles for uPA, with the control and GDF-
Observed from both of the 9 treated samples. 18-20 cycles were determined to be optimal and used to study all three gene products in all further analyses. Similar PCR conditions were used for COX-2, LH receptor, P-450scc, and S
A tAR study was performed (see below).

Next, the dose-response relationship between recombinant GDF-9 and HAS2 mRNA synthesis was examined. The rate of hyaluronic acid synthesis by cumulus cells or mural granulosa cells exposed to the oocyte conditioned medium peaked at 6-12 hours in culture (sarstri (
Salustri, A.) et al., J. Biol. Chem., 264: 13840-13847 (1989) and D'Alessandris, TE et al., J. Biol. Chem., 272: 4787-4794 (1997). ). Thus, for dose response experiments, granulosa cells were cultured in the absence or presence of various concentrations of recombinant GDF-9 for 5 hours. 10 ng using semi-quantitative RT-PCR
A slight increase in HAS2 expression can be detected with / ml recombinant GDF-9. The level of recombinant GDF-9 between 30-300 ng / ml is 3
Strong HAS with relatively linear dose / response occurring between 0-120 ng / ml
Two leads were given. Granulosa cells recovered from unprimed immature mouse ovaries also responded to recombinant GDF-9. In contrast, recombinant mouse BMP-
15 (150 ng / ml) or recombinant human BMP-6 (50 ng / ml) failed to stimulate HAS2 expression or suppress uPA expression when tested in any of the granulosa cell assays. These findings indicate that their activity was GDF-9.
And show that other oocyte-secreted TGF-β family members are unable to respond to these activities.

Time course analysis of HAS2 expression pattern using 100 ng / ml GDF-9 (0-24 hours) showed slight induction of HAS2 mRNA in 2 hours,
Peak induction was seen during 5 hours, and HAS2 expression levels were shown to decrease by 9 hours in culture. Granulosa cells cultured in the absence of GDF-9 for 0 to 24 hours
Expresses more than 10-fold lower levels of HAS2 compared to DF-9-induced peak levels, indicating very low levels of basal activity in these cells. The time course for uPA expression in control granulosa cells cultured in the absence of GDF-9 showed that uPA levels increased over the first 9 hours in culture. However, granulosa cells treated with 100 ng / ml recombinant GDF-9 maintained detectable but very low levels of uPA expression. 5 hours, GD
The band intensity from the F-9 treated sample was approximately 15% of the control and at 9 hours was approximately 9% of the band intensity of the control treated sample (normalized to HPRT for each sample).

GDF on HAS2 and uPA expression as shown by RT-PCR
To confirm the effect of -9, the expression of HAS2 in primary granulosa cells was examined by Northern blot analysis in the presence or absence of 50 ng / ml GDF-9. Total RNA from each sample (0 or 5 hour incubation in the presence or absence of 50 ng / ml GDF-9) was subjected to Northern blot analysis and hybridized with either the HAS2 or uPA probe and then with the GAPDH probe. Let it soy. The signal was quantified on a phosphor imager for HAS2 and uPA levels and normalized to GAPDH. HAS2 was barely detectable in mural granulosa cells at 0 hours or 5 hours in culture in control samples. However, 4.8 kb after 5 hours of incubation with GDF-9.
And 3.2 kb of the HAS2 mRNA type (Spicer, AP) et al.
Biol. Chem., 271: 23400-23406 (1996)) increased 9.7-fold compared to the control. In contrast, Northern blot analysis of uPA showed that 50 ng / ml GDF-9
It was shown that A synthesis was suppressed to 40% of the control medium. Thus, Northern blot data confirmed our RT-PCT analysis.

Recombinant GDF-9 Causes Cumulus Expansion of Oocyte-Resected Cumulus Cells-Oocyte Complexes From Immature Female Mice Treated With Untreated Cumulus Cell-Oocyte Complexes Isolated. Using the set transgenic micromanipulator,
Oocytes from these complexes were punctured and the contents of the oocytes were aspirated. Prior to culture, these oocyte-excised cumulus complexes were spheroids approximately 100 μm in diameter consisting of several layers of granulosa cells surrounding the empty zona pellucida. After 18 hours of culture, control medium (ie, GDF-9 deficient but 10% fetal calf serum and 5%
Cumulus cells from 25 of the 25 oocyte resection complexes (containing ng / ml or 150 ng / ml FSH) adhere to tissue culture plates and have the appearance of fibroblasts. . Consistent with previous reports that cumulus cells have low or undetectable levels of LH receptor mRNA, these oocyte ablation complexes (9 out of 9) with LH (1 μg / ml) ) Incubation did not alter the appearance of the fibroblasts. In contrast, isolated under the same conditions,
Forty of the forty-four oocyte resection complexes cultured in the presence of 00 ng / ml GDF-9 maintained a spherical appearance and expanded into three-dimensional gelatinous spheres.
These results are similar to the cumulus cell-oocyte complex with untreated oocytes cultured in FSH containing media. These cells did not detach from the plate due to cell death, as the majority of cells continued to exclude the vital dye trypan blue. In contrast, incubation of the complex with recombinant human BMP-6 or BMP-15 did not result in cumulus expansion. These observations indicate that GDF-9 specifically stimulates cumulus expansion and is an oocyte-derived factor that normally mediates this process.

Regulation of LH receptor and cyclooxygenase 2 (COX-2) by GDF-9 Based on the above findings, LH receptor and COX2 in granulosa cell culture medium
The effect of recombinant mouse GDF-9 on expression was determined using semi-quantitative RT-PCR. After collection of granulosa cells (T = 0), low levels of LH receptor were detected. In the absence of GDF-9, LH receptor levels were further reduced, but expression increased between 5 and 24 hours. In contrast, GDF-9 (100 ng / ml
)), The LH receptor m
RNA synthesis was suppressed. LH receptor mRNA was suppressed in the presence of 100 ng / ml GDF-9, even in the presence of 10 ng / ml FSH. 10ng / m
In experiments shown with 1 FSH, control-treated granulosa cells were
-9 expressed about 30 times more LH receptor than granulosa cells treated (p <0.0
5, n = 3). In contrast, COX-2 expression was low in the absence of GDF-9, but showed dramatic elevated levels after 24 hours of incubation in the presence of 100 ng / ml GDF-9. FSH had no significant effect on either control or GDF-9 treated granulosa cells, while GDF-9 caused a more than 50-fold increase in COX-2 expression. Analysis of HAS2 expression in these samples was consistent with previous results.

GDF-9 Modulation of Progesterone Synthesis Dose response and time course experiments were performed to study the effects of GDF-9 and FSH on the regulation of progesterone synthesis. Perform semi-quantitative RT-PCR and
-450scc mRNA synthesis and StAR mRNA synthesis were analyzed, and progesterone secreted into the medium was analyzed using a radioimmunoassay. 50n
Media was collected from granulosa cells cultured for 4 hours in triplicate wells containing various amounts of FSH in the presence or absence of g / ml GDF-9 and assayed for progesterone by radioimmunoassay. Under these conditions, granulosa cells cultured for 4 hours in the absence or low levels of FSH produced very little progesterone. On the other hand, granulosa cells cultured with 50 ng / ml GDF-9 produced 5-7 times higher amounts of progesterone (Figure 4A). Granulosa cells were incubated with medium containing 10% FCS with 0 or 10 ng / ml FSH in the presence or absence of GDF-9 for 24 hours. However, 5 ng / m
The amount of progesterone produced in the presence of 1 FSH is not significantly affected by the presence or absence of GDF-9, but in the presence of GDF-9 alone or low levels of FS
About 2-3 times more than that produced in H. After 24 hours of incubation, only GDF-9 induced significantly higher amounts of progesterone, but higher concentrations of FS
In the presence of H (10 ng / ml), the effect of GDF-9 was negligible (
(Figure 4B). These data indicate that stimulation of progesterone FSH and GDF-9 is not additive.

To determine how FSH and GDF-9 modulate progesterone synthesis in granulosa cell cultures, in the presence or absence of FSH or GDF
-Quantitative RT- after 24-hour incubation in the presence or absence of -9
The levels of P-450scc and StAR mRNA were analyzed by PCR (Fi
gure4C-4D). On the other hand, 10 ng / ml FSH caused a slight increase in P-450scc and the presence of GDF-9 had no effect. In contrast, GDF-9 resulted in a 2- to 5-fold induction of StAR mRNA levels (p <
0.05). These results indicate that FSH and GDF-9 act through different mechanisms to regulate key progesterone synthesis.

Discussion Using immunohistochemical analysis, the GDF-9 protein was first detected at low levels in mature oocytes of the primary follicle (type 3a follicle), and the type 3b follicle fully matured oocyte It is shown herein to be present at higher levels in cells and to be detected in oocytes at each subsequent developmental stage. Furthermore, GDF- is also produced by the oocytes of the follicles of the large follicle and the preovulatory follicle in which the oocyte is closely associated with the granulosa cell.
Nine proteins have been shown to be synthesized. GDF-9 protein is 3
It is detected in oocytes of type a follicles, but it is only essential for follicle formation in type 3b-4 type follicular translocation. This indicates that the GDF-9 signaling cascade is not active before the type 3b follicular stage.

As shown herein, GDF-9 replaces oocyte and oocyte conditioned media in assays analyzing HAS2 induction and uPA suppression typical of processes occurring in preovulatory follicles. Obtained. As also shown herein, other oocyte-expressed TGF-β family members, BMP-15 and BMP-6, could not substitute for GDF-9 in these granulosa cell assays. Mural granulosa cells isolated from follicular cavity follicles treated with recombinant GDF-9 were induced to express HAS2 in a dose-dependent and time-dependent manner. GDF-9 induced about 10-fold higher levels of HAS2 mRNA in mural granulosa cells, which corresponded well to the maximal effect of oocytes on HA synthesis. Furthermore, the dose response curve for GDF-9 was very similar to that of the oocyte conditioned medium. Very low doses (eg, 0.5 oocytes / μl or 10 ng / ml GDF-9) induce very low but detectable increases in hyaluronic acid synthesis or HAS2 expression, while one oocyte Cells / μl or 30-50 ng / ml GDF-9 caused a greater and more dramatic induction to reach a stable state with 2-4 oocytes / μl or 120-300 ng / ml GDF-9 (sarstri ( Salustri, A.) et al., Dev. Biol., 138.
: 26-32 (1990)). The time course of GDF-9 action was also consistent with previous data on oocyte production factors. On the other hand, HAS2 mRNA was induced by 2 hours of culture using GDF-9, and oocyte-induced hyaluronic acid became detectable at a low level after 2.5 hours (Sarstri et al., J. Biol. Chem., 264: 13840-13847 (1
989)). GDF-9 induced peak HAS2 mRNA levels between 3-5 hours, while the rate of oocyte-induced hyaluronic acid synthesis was maximal between 6-12 hours (Sarstri et al., J. Biol. Chem., 264: 13840-13847 (1989)). Similarly,
Oocyte-induced hyaluronic acid synthesis was reduced after 12 hours and no more hyaluronic acid was produced after 18 hours (D'Alessandris, TE)
J. Biol. Chem., 272: 4787-4794 (1997)); GDF-9 induced HAS2 expression
By 24 hours, uPA synthesis had increased by 24 hours. The "transient" nature of HAS2 expression and activation of hyaluronic acid synthesis and increased uPA synthesis during this period is due to the instability of GDF-9 in the medium, down-regulation of GDF-9 receptor, granulosa cells GDF-9-induced differentiation of and / or GDF-
It appears to be due to stimulation of a negative feedback mechanism in the 9 signaling cascade. The oocyte secreted factors that regulate some of these processes have been found to be unstable, and various co-cultures containing granulosa cells (co-cu
It is interesting to note that the continued presence of oocytes in the ltures is required for continued activity (Eppig, JJ et al., Biol. Reprod., 56: 9
76-984 (1997)). Finally, the data described herein show that although the differences in in vivo phenotype between mural and expanding cumulus granulosa cells are not unique to the cells themselves, their egg This shows that it is due to the proximity to the mother cell and the concentration gradient of GDF-9 produced by the oocyte.

The conversion of cholesterol to pregnenolone is a kinetic step in granulosa cell steroidogenesis. The rate of pregnenolone synthesis depends on the level and activity of the enzyme cytochrome P-450 side chain cleavage (P-450scc), which catalyzes the reaction, and access to the substrate cholesterol through stimulation of the steroidogenic early regulatory protein (StAR). (Rennert, H. et al., "Intracellular cholest
erol dynamics in steroidogenic cells ", In: Adashi (LC Adashi, EY) et al. (
Ed.) The Ovary, Raven Press, Ltd., New York, 147016
P. 4 (1993)). The increase in FSH- and LH-induced intracellular cAMP is
It is well established that it leads to subsequent stimulation of P-450scc, StAR mRNA synthesis and StAR protein phosphorylation, and stimulates in vitro progesterone synthesis (Richards, JS and Hedin, L., Annu. Rev.). . Phy
siol., 50: 441-463 (1988); Clark, BJ et al., Mol. Endocr., 9: 1346-55.
(1995) and Arakane, F. et al., J. Biol. Chem., 272: 32656-62 (1997)).
In contrast to the effects seen with GDF-9 treatment, activin A was found to be basal and FSH stimulated P-450scc, 3β hydroxysteroid dehydrogenase (3βHSD
) And brogeslone synthesis were reduced by cultured granulosa cells from rats stimulated with diethylstilbestrol (DES). The data described herein is based on FS
Although H was confirmed to stimulate P-450scc mRNA synthesis in mouse granulosa cells, it indicates that GDF-9 did not significantly affect P-450scc mRNA synthesis. In contrast, FSH has only a small inductive effect on StAR mRNA, whereas GDF-9 with or without FSH significantly induces StAR expression. As a result, both GDF-9 and FSH are able to increase progesterone production by granulosa cells independently, and appear to act through the same pathway but different mechanisms. This local production of progesterone by cumulus cells appears to be essential in achieving a complete microenvironment for oocytes after ovulation and before fertilization. Evidence of this is that the ovulated rat cumulus cell-oocyte complex secretes measurable levels of both progesterone and prostaglandins (primarily PGE2) (Schuetz, AW). And Dubin, NH, Endocr., 108: 457-463 (1981))
And the use of aminoglutethimide, which inhibits the conversion of cholesterol to pregnenolone, reduces the number of normal sheep oocytes recovered after in vitro follicle maturation.

As described herein, in situ hybridization analysis of LH receptors in preovulatory follicles indicates that COX2 expression is highest in cumulus cells after LH treatment in vivo, whereas LH receptors are It shows that it is suppressed in cells, but not in mural granulosa cells. As described herein, recombinant GDF-9 suppresses LH receptor mRNA but induces COX-2 expression, mimicking the normal expression of these genes in cumulus cells. . However, Eppig and his colleagues (Eppig, JJ)
Biol. Reprod., 56: 976-984 (1997) and Eppig et al., Mol. Reprod. Dev., 49.
: 327-332 (1998)) states that fully mature oocytes suppress LH receptor mRNA expression, but oocytes derived from prefollicular cavity follicles, oocytes arrested in metaphase II, or two-cell embryos It was briefly indicated that it was not valid. Knockout Research (Dong, J
., Et al., Nature, 383: 531-535 (1996) and Elvin (JA) et al., Mol. Endoc.
r. (submitted) (1999)) and, as described herein, GDF-9.
Can stimulate changes in cell morphology, gene expression, and steroidogenesis, and granulosa cells from at least primary follicles and follicular cavity follicles are GDF-
9 (see Figure 5 for a summary of the findings described herein). Because GDF-9 dramatically increases the level of COX-2 expression, either COX-2 expression is not expressed at an earlier stage of follicle formation or an earlier stage of oocytes (Eppig, JJ et al.) , 19
97 Biol. Reprod., 56: 976-984 (1997) and Eppig et al., Human Reprod., 12: 127.
-132 (1997) and Eppig et al., Mol. Reprod. Dev., 49: 327-332 (1998)) are not clear why they are less effective at inhibiting the LH receptor. GDF-9 protein is detected at the immunohistochemical level at all stages of follicle formation, but this does not necessarily prove that GDF-9 is active at all these stages. One possibility is that the GDF-9 precursor is only processed into an active mature dimer at the 3b type stage and the follicular cavity follicle stage. Regulation of the GDF-9 processing enzyme may be one way to regulate the activity of GDF-9 post-translationally. Another explanation is that both GDF-9 regulated and GDF-9 independent transcription factors are COX-2
Act together to regulate the synthesis of. At least one regulator of COX-2 is a transcription factor enhancer binding protein P (C / E
BPβ). C / EBPβ, which is induced for 4-7 hours after hCG treatment, binds to the COX-2 promoter and down regulates COX-2 mRNA expression. COX-2 expression in the ovaries usually peaks 4 hours after hCG treatment, while COX-2 protein remains present in cumulus cells after ovulation (Lim, H. et al., Cell, 91 : 197-208 (1997)). However, C / EBPβ knockout mice (which lack fertility [Sterneck, 1997 # 279])
In, the levels of COX-2 mRNA remain elevated. COX-
2 knockout mice are infertile due to defective ovulation and impaired oocyte maturation (Lim, H. et al., Cell, 91: 197-208 (1997)). These studies from the C / EBPβ and COX-2 knockout models indicate that proper regulation of COX-2 in cumulus cells is necessary to maintain the optimal microenvironment of prostaglandins around oocytes. Is shown. GDF-9 appears to be one of the factors involved in COX-2 induction, and C / EBPβ plays a role in COX-2 downregulation. The work described herein and others (Lawrence, T. et al., Endocr., 106: 1114-1118 (1980) and Meduri, G.) et al., Endocr., 131: 36.
6-373 (1992)) also show that LH indirectly regulates expression of both COX-2 and C / EBPβ in cumulus cells, since the LH receptor is not present on these cells. I have. However, it is not clear how this is achieved or how COX-2 appears rapidly and C / EBPβ appears more slowly.

Example 4 Molecular Characterization of Follicular Defects in Growth Differentiation Factor 9-Deficient Ovaries Materials and Methods Experimental Animals All experimental mice were cared for and used by laboratory animals.
ry Animals). Unless otherwise specified, 6
Ovaries from Ｃ12-week-old adult C57B1 / 6 / 129SvEv hybrid mice were used for both RNA isolation and preparation for in situ hybridization and immunohistochemistry. For COX-2 experiments, 7.5 IU of Gestil (Diosyns, BV, The Netherlands) was given to 3-week old mice, followed by 48
After time, 5 IU of hCG (Sigma, St. Louis, Mo.) was injected intraperitoneally. Ovaries were harvested 5 hours after hCGu injection.

RNA Isolation and Northern Blot Analysis Total RNA was obtained from various tissues of wild-type and GDF-9 deficient C57B1 / 6 / 129SvEv hybrid mice, RNA STAT-60 (Reed Medical Laboratories, Houston, Tex.). Was extracted as described by the manufacturer and quantified on a spectrophotometer. 15 μg of total RNA for each sample
Electrophoresis on a 1.2% agarose / 7.6% formaldehyde gel, Hyb
and N nylon membranes (Amersham, Arlington Heights, Ill.). Table 2 provides a summary of all specific cDNA fragments used to generate the probes in these experiments. Strip-EZ probe synthesis kit (
Ambion, Austin, Tex. Using {α- ³² P} dAT
A probe was prepared by random priming with P. The membrane was hybridized, washed and described (Mahmoudi, M. and Lin, VK, Biotech., 7: 331-332 (1
Autoradiography was performed as in 989)). Strip-EZ removal reagent (
The probe was removed from the membrane using Ambion) according to the manufacturer's protocol.
The same blot was then reprobed with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or 18S ribosomal RNA as a loading control.
(reprobe) The signal of each probe was quantified with a molecular dynamics optical densitometer.

[0078]

[Table 2]

In Situ Hybridization Essentially described above with the following changes (Albrecht, U., et al., “Visualization
of gene expression patterns by in situ hybridization ", In: GP Daston
(ed) Mol. Cell. Meth. Dev. Toxic., CRC Press, Boca Raton, FL, pp. 23-48
, (1997)). Freshly excised ovaries from wild-type or GDF-9 deficient C57B1 / 6 / 129SvEv hybrid mice were fixed in 4% paraformaldehyde-PBS overnight,
Treated and embedded in paraffin. 5 μm thick sections were cut and pretreated as described. Table 2 contains a summary of specific probe information. {Α- ³⁵ S} UTP
-Labeled antisense and sense probes were replaced with Riboprobe
) Made using the T7 / T3 or Riboprobe T7 / SP6 combination system (Promega, Madison, Wis.). Using 5 × 10 ⁶ cpm of each riboprobe per slide, 50% deionized formaldehyde / 0.3 M NaCl / 20 mM TrisHCl (pH 8.0) / 5
Hybridization was performed at 55 ° C. for 16-18 hours in mM EDTA / 10 mM NaPO ₄ (pH 8.0) / 10% dextran sulfate / 1 × Denhardt's solution / 0.5 mg / ml yeast RNA. A high stringency wash of 2 × SSC / 50% formamide and 0.1 × SSSC at 65 ° C. was performed. Dehydrated sections were immersed in NTB-2 emulsion (Eastman Kodak, Rochester, NY) and exposed at 4 ° C for 2-14 days depending on the probe. After development, slides were counterstained with hematoxylin and mounted for photography.

The ovaries were fixed in 20% neutral buffered formalin for 3 hours, processed, embedded in paraffin and minced to a thickness of 4 μm. (Robker, RL, and Richards, JS, Mol.
Endocr., 12: 924-940 (1998)), and PCNA was detected using a mouse anti-PCNA monoclonal antibody (Novocastra Laboratories, Newcastle upon Tyne, UK). . Rabbit anti-mouse P-450scc polyclonal antiserum was obtained from Michael J. of the University of Kansas Medical Center. Courtesy of Michael J. Soares, 2% normal mouse serum (Sigma) and 2% normal goat serum (Vector Laboratories
), 1 × PBS, including Burlingame, Georgia), 0.0
Used at a 1: 625 dilution in 5% Tween 20 (PBST). 1% containing 2% normal mouse serum and 2% normal goat serum BSA, in 0.1% NaN ₃ in PBS, rabbit anti-Ki-67 polyclonal antiserum (Nobokasutora) 1: 300
And a 1: 125 dilution of rabbit anti-p27 polyclonal antiserum (Santa Cruz Biotechnology, Inc., Santa Cruz, Calif.). Good staining for both Ki67 and p27
An antigen retrieval method was required. For Ki-67 staining, sections were cooked in 0.1 M citrate buffer pH 6.0 for 35 minutes. For p27 staining, sections were cooked for 35 minutes in a Tris-EDTA antigen recovery solution at pH 8.0. For Ki-67, p27 and P-450scc, all sections were
. Blocked in 1 × PBS containing 05% Tween 20, 2% normal mouse serum (Sigma) and 2% normal goat serum for 30 minutes and incubated in primary antibody for 1 hour at room temperature. Using a Super Sensitive Mouse Antibody Animal Detection Kit (Biogenex, San Ramon, CA) containing an anti-mouse IgG biotinylated secondary antibody preabsorbed in rat tissues , PCNA detection. P-450scc, p27 and Ki-67 antibodies were detected using a Super Sensitive Rabbit Antibody Detection Kit (Biogenex) containing an anti-rabbit IgG biotinylated secondary antibody preabsorbed in mouse tissues. PCN
A and P-450scc are detected using an alkaline phosphatase label conjugated to streptavidin and a new fuchsin substrate (Biogenes), while p27 is a horseradish peroxidase label conjugated to streptavidin (Biogenes) and Detection was performed using DAB substrate (Vector Laboratories).

TUNEL Assay Apoptag Plus Complete Apoptosis Detection (Apopta
g Plus Complete Apoptosis Detection) Kit (Oncol Laboratories (Onc
The ovaries were stained for apoptotic cells by a modified TUNEL method using the method described in G.I. The nuclei were counterstained with a medium containing Propidium iodide / Antifade (Oncol Laboratories).

RT-PCR Analysis cDNA-primed with oligo-dT from 1 μg of either control or GDF-9 deficient ovary RNA was ligated to Superscript reverse transcriptase (Gibco BR
L) according to the manufacturer's instructions. 1 μl of each RT reaction (
Kit ligand specific oligonucleotide spanning an intron and an 84 bp alternatively spliced exon: 5 'CCAGAAACTAGATCCT
TTACTCCT 3 '(SEQ ID NO: 17) (nts493-51 of sense-S40364)
7) 25 each of which is primed with a primer and a 5′CTGTTGCAGCCAGCTCCCTTAG 3 ′ (SEQ ID NO: 18) (943 セ ン ス 919 of antisense S40364) primer
Used in μl PCR reactions. KL-1 type amplification yields a 450 bp product and KL-2 type amplification yields a 366 bp product. The products were separated on a 2% agarose gel and visualized by ethidium bromide staining.

Results Cell Cycle Progression GDF-9 deficient follicles with untreated oocytes contain only one granulosa cell and do not form a follicle consisting of multiple layers of granulosa cells. Proliferating cell nuclear antigen (PCNA), a cofactor for DNA polymerase δ and the cyclin-cdk complex, is expressed in G1, increased through the G1 / S transition, high in G2, and sharply decreases in the M phase of the cell cycle. (Oktay, K. et al., Biol. Reprod., 53: 295-301 (1995)). Similarly, Ki-67, a component of the granulosa nucleolus, is expressed in all phases of the cell cycle except G0 (Gerdes, J., et al., J. Immunol., 133: 1710-1715). (1984)
). The most proliferative granulosa cells are the majority of granulosa cells with PCNA and Ki-
It is found in the 67-positive wild-type follicular cavity. Wild type ovary type 3b (large 1
In layer) and type 4 (layer 2) follicles, immunohistochemical analysis of PCNA or Ki67 revealed that> 50% of granulosa cells at the section were positive (ie, positive was defined as strong red staining of the nucleus and negative) Is defined as pale or diffuse staining of the nucleus and cytoplasm). In contrast, type 3b follicles with untreated oocytes in GDF-9 deficient ovaries show <10% positively stained granulosa cells / section, which indicates that almost all granulosa cells are blocked by GO. It was shown to be. However, shortly after oocyte degeneration, granulosa cell differentiation began to occur in the follicles of GDF-9 deficient ovaries and more cell nuclei stained positive for both PCNA and Ki-67. Oocyte nucleus staining using PCNA antibody is GDF-9 deficient 3a
And 3b were detected in the mature and fully mature oocytes of follicles, indicating previous studies showing the presence of PCNA in the nuclei of oocytes from primary to follicular follicles in wild-type ovaries (Oktay, K. et al., Biol. Reprod., 53: 295-301
(1995)).

[0086] Normal follicle development in TUNEL-labeled GDF-9 deficient ovaries ceases at the 3b type stage and oocyte loss ultimately results in granulosa cell "differentiation", but histology of GDF-9 deficient ovaries Surprisingly, the apoptotic presentation (appearin
g) No cells were detected. To confirm the relative absence of apoptotic cells in GDF-9 deficient ovaries, a TUNEL assay was used to fluorescently label DNA ends. Granulosa cells and the corpus luteum of the wild-type regression follicle follicles contained a large number of positively labeled cells. In contrast, sections of GDF-9 deficient ovaries were 2 per ovarian section.
Contains 10 to 10 positively labeled cells. No positively stained cells were found in monolayer follicles, but occasionally apoptotic cells were found in steroidogenic follicular nests or stromal tissue in follicles with degenerating oocytes. Therefore,
In the absence of GDF-9, once the type 3b follicle stage is reached, the majority of granulosa cells remain dormant and do not proliferate or die.

Analysis of Cell Cycle Inhibitors p21 and p27 are well documented cell cycle inhibitors and are involved in cell cycle arrest during luteal formation in the ovaries (Robker, RL and Richa
rds, JS, Biol. Reprod., 59: 476-482 (1998)). Based on the relative lack of proliferation of granulosa cells in GDF-9 knockout ovaries, the expression of both p21 and p27 mRNA was examined by p27 protein by in situ hybridization and immunohistochemistry. p21 mRNA is wild-type and GDF
Low levels are ubiquitously detected in both -9 deficient ovaries, and high levels are detected in wild type regressive follicles and scattered cells in the corpus luteum and luteinized follicles of GDF-9 deficient ovaries Was done. In the wild-type ovary, p27 m
RNA was ubiquitously expressed at low levels throughout the ovaries, but was more abundant in the corpus luteum of wild-type ovaries. In GDF-9 deficient ovaries, monolayer follicular granulosa cells expressed detectable levels of p27 message, and cells in the middle subgroup of GDF-9 deficient ovaries expressed higher levels. Similarly, nuclear p27 immunoreactivity was clearly detectable in the majority of luteinizing granulosa cells in the wild-type luteum, as well as in the luteinizing follicles of GDF-9 deficient ovaries. Also, a decrease in p27 nuclear staining was in granulosa cells of both wild-type and GDF-9-deficient monolayer follicles, which was clearly stronger than in negative controls or stromal cells. Although it is not possible to estimate protein levels by immunohistochemistry, there appears to be no dramatic difference in p27 immunoreactivity between monolayer follicles of GDF-9 knockout ovaries and wild-type ovaries. It is. This indicates that p27 protein overexpression is not the reason for blocking follicle formation at the primary follicular stage in GDF-9 deficient ovaries.

It has previously been reported that morphologically distinct follicular membrane layers could not be detected by light and electron microscopic analysis in GDF-9 deficient ovaries (Do
ng, J., et al., Nature 383: 531-535 (1996)). However, the flattened layer of fibroblasts outside the granulosa cell basement membrane surrounds the type 3b follicle circularly in GDF-9 deficient ovaries. 17α-hydroxylase cytochrome P-450 (17α-OH), which is a follicular membrane cell-specific enzyme required for androgen production
In-situ hybridization was performed with the probe for (1), and it was confirmed that the follicular membrane layer was actually present. In the wild-type ovary, cells expressing 17α-OH associate into type 3b and type 4 follicles and form a complete ring just outside the granulosa cell basement membrane by the multilamellar prefollicular cavity follicle stage. Began to do. In contrast, in GDF-9 deficient ovaries, only a few cells expressing 17α-OH are present and scattered in stromal cells and do not collect in follicles. These 17α-OH positive cells appear to be a group of follicular membrane cell precursors that respond to elevated serum LH (Do
ng, J., et al., Nature 383: 531-535 (1996)). Similarly, the absence of the LH receptor and c-kit mRNA around the follicle (see below) confirms that no follicle membrane cell layer is formed. These data indicate that GDF-9 signaling is required, either directly or indirectly, to recruit follicular cell precursors to early prefollicular follicles.

Analysis of c-kit and Kit Ligand Expression The c-kit / kit ligand signaling pathway has been shown to be important for germ cell proliferation and follicle formation (Besmer, P., et al., Dev. Suppl. ., 125-137 (1993)
; Huang, EJet al., Dev. Biol., 157: 100-109 (1993) and Kuroda, H., et al., Dev.B
iol., 126: 71-79 (1988)). Northern blot analysis revealed that c-kit mRNA
Was expressed in GDF-9 deficient ovaries, and levels were shown to be comparable or slightly higher than wild-type ovaries. By in situ hybridization, c-kit mRNA was localized to oocytes and follicle-stromal cells of wild-type ovaries, but was previously shown (Manova, K., et al., Devel., 110 1057-1069 (1990)). In GDF-9 deficient ovaries, c-kit
The mRNA was localized only in the oocytes, with slight background levels of silver grains present in other cell lines.

In contrast to the results of c-kit expression, kit ligand expression in GDF-9 deficient ovaries was increased 32 fold compared to expression in wild type ovaries. IGF-I
It is also expressed in granulosa cells of early pre-follicular follicles, but does not show a similar increase in expression in GDF-9 deficient ovaries, which indicates that increased kit ligand is associated with tissue composition. Indicates that they exhibited specific regulatory interactions rather than effects. Although in situ hybridization is not a reliable method for quantifying mRNA expression, the relative expression level between wild-type and GDF-9 deficient ovaries depends on the hybridization efficiency and between slides in emulsion concentration. To minimize variability, comparisons can be made by positioning sections from both types of ovaries in close proximity to each other on the same slide. Under these conditions, kit ligand expression is
Barely detected in wild-type ovaries, a faint signal above background appeared in granulosa cells of prealveolar follicles. In contrast, in the GDF-9 deficient ovaries, the expression of the kit ligand in granulosa cells was abundant. Type 3a and early type 3b follicles have detectable levels of kit ligand, while the largest monolayer follicles showed stronger staining. If paracrine factors produced by the multilayered follicles of the adult ovary suppressed kit ligand expression, ovaries from 10, 17 and 28 day old wild type mice were examined. Although the kit ligand was somewhat easier to detect in these immature ovaries due to an increase in the number of prefollicular follicles, the relative expression level per follicle was similar to GDF-9 deficiency at similar ages. It was not equivalent to the levels found in the ovaries. Kit ligand expression was further increased in granulosa cells within asymmetric follicles, presumably fate of undergoing oocyte degeneration. However, shortly after the oocyte was denatured, kit ligand expression disappeared and was not present in the follicular nest.

There are two alternatively spliced forms of KL, KL-1 and KL-2, which differ by 84 bp. This alternate splicing results in the addition of 28 amino acids in KL-1, which contains the proteolytic site. The membrane-bound KL is
The cell-bound form is consequently more effective because it is more active and more stable than the released KL, KL-2 (Besmer, P., et al., Dev. Suppl., 125-137 (1993). )).
Non-quantitative RT-PCR using primers that could distinguish KL-1 from KL-2 detected both forms of KL in both wild-type and GDF-9 deficient ovaries. Immunohistochemical analysis of KL protein in GDF-9 deficient ovaries showed the same expression pattern as KL mRNA and resulted in the strongest staining in asymmetric follicles. Taken together, these results indicate that GDF-9 negatively regulates kit ligand expression in granulosa cells in a paracrine fashion and that active K
2 shows that L protein is synthesized.

TGF-β superfamily members (activin, inhibin, follistatin) Activin and inhibin have been implicated in the regulation of granulosa cell proliferation and follicle maturation both in vivo and in vitro (Mather, JP, et al. , Proc.Soc.Exp
.Biol.Med, 215: 209-222 (1997) and Matzuk, MM, et al., Rec. Prog. Horm. Res., 51
: 123-157 (1996)). The expression levels and patterns of inhibin α, activin βA, activin βB and the activin binding protein follistatin were examined in wild-type and GDF-9 deficient ovaries. In Northern blot analysis, surprisingly, inhibin alpha was expressed at similar levels in GDF-9 deficient ovaries and wild-type ovaries. In wild-type ovaries, inhibin alpha was expressed in granulosa cells of all mature follicles (from type 3a to preovulatory) but was excluded from the corpus luteum. In GDF-9 deficient ovaries, inhibin alpha was highly expressed in monolamellar follicles, follicles with degenerating oocytes, and central steroid-producing ovaries. This is because granulosa cells in wild-type ovaries are similar to the corpus luteum in many respects, but proliferate, form multilamellar follicles, and assemble into the active follicular membrane layer before lutealization. Indicates that they are developmentally different.

The two inhibin / activin β subunits, βA and βB, and the activin binding protein, follistin, were expressed in overlapping patterns in wild-type ovaries. All three genes were expressed in multilayered prefollicular follicles and follicular granulosa cells. Interestingly, expression of βB in regressive follicles was low and specifically localized around the oocyte, but high levels of all granulosa cells were expressed in healthy follicles. GDF-9 deficient ovaries have very low levels of βB message and follistatin, which are almost undetectable by Northern blot analysis, which are weakly localized in granulosa cells of monolayer follicles; It was more robustly localized in follicles with certain oocytes; and at the highest level in oocyte-deficient follicular nests. The βA message was not detectable in monolayer follicles, but was weakly localized in granulosa cells of follicles with degenerating oocytes and was expressed at high levels in oocyte-deficient follicular nests. However,
By Northern blot analysis, βA expression levels were comparable between GDF-9 deficient ovaries and wild-type ovaries. These results indicate that GDF-9 deficient ovaries maintained the ability to make both activin-β and inhibin α subunits.

Characterization of Follicle Cavity Follicle Markers (ERβ, FSHR, Aromatase Cytochrome P450) To further characterize the follicles of GDF-9 deficient ovaries, other markers associated with granulosa cell space functional differentiation: FSH Receptor (FSHR), aromatase cytochrome P-450 (aromatase) and estrogen receptor β (
ERβ) expression was analyzed. Previously, Northern blot analysis showed that FSHR was similarly expressed in both GDF-9-deficient and wild-type ovaries, and that aromatase expression was 50% of wild-type levels (Dong, J., et
al., Nature, 383: 531-535 (1996)). ERβ was expressed at low but similar levels in knockout ovary and wild-type ovary. Previously reported by in situ hybridization in wild-type ovaries (Richards, JS, Endocrine R
eviews, 15: 725-751 (1994); Byers, M., et al., Mol. Endocr., 11: 172-182 (1997) an
d Camp T., et al., Mol. Endocr., 5: 1405-1417 (1991)), FSHR,
ERβ and cytochrome aromatase were specifically detected in granulosa cells.
ERβ was expressed at low levels in unilamellar follicles and at higher levels in multilamellar follicles. FSHR was expressed in multilamellar prefollicular and follicular follicles. Aromatase is normally induced by FSH stimulation of granulosa cells, but was specifically expressed at high levels in preovulatory follicles. In GDF-9 deficient ovaries, ERβ
Was expressed at low levels in monolayer follicles and at somewhat higher levels in follicles with degenerating oocytes. Also, FSHR was expressed in follicles with degenerating oocytes, while aromatase was expressed at high levels only in follicle nests with fully degenerated oocytes. Aromatase expression in these follicles in GDF-9 deficient ovaries indicates that a functional FSH signaling pathway similar to the granulosa cells of preovulatory follicles of wild-type ovaries is present.

Characterization of Preovulatory Markers and Luteal Markers (COX-2, LHR, Cholesterol Side Chain-Cleaved Cytochrome P-450) GDF-9 Deficiency with the Appearance of Lutealizing Granulosa Cells as Described Previously The ovary contains multi-layered, centrally located nests. By electron microscopy analysis, these nests contain a large number of lipid droplets and mitochondria with tubular crystals that are highly typical of steroidogenic cells (Dong, J., et al., Nature, 383: 531-535 ( 1996)). To confirm their steroidogenesis and similarity to luteal cells, we analyzed the expression of cyclooxygenase 2 (COX-2) and LH receptor (LHR) messages, and the expression of cholesterol side chain-cleaved cytochrome P-450 protein (P-450scc). did. COX-2 is expressed in wild-type ovaries only at discrete times during the period following LH surge (Sirois, J., et al., J. Biol. Chem., 267: 11586-11592 (
1992) and Sterrieck, E., et al., Genes Dev., 11: 2153-2162 (1997)), PMSG
Ovaries from 3-week-old mice stimulated for 48 hours and then hCG for 5 hours were used for both Northern blot analysis and in situ hybridization.
By Northern blot analysis, RNA from superovulated ovaries showed three distinct bands, but no bands could be detected in lanes of unstimulated wild type or GDF-9 deficient ovaries. In agreement with the Northern blot data, in situ hybridization showed that COX-2 was expressed by granulosa cells of preovulatory follicles in wild type ovary stimulated with PMSG and hCG (
Sterrieck, E., et al., Genes Dev., 11: 2153-2162 (1997)). The highest expression occurred in cumulus cells of wild type follicles at 5 hours, but COX-2 expression was completely undetectable in GDF-9 deficient ovaries.

In contrast to the COX-2 expression data described above, Northern blot analysis has previously shown that LHR is expressed in GDF-9 deficient ovaries at levels comparable to wild-type ovaries (Dong , J., et al., Nature, 383: 531-535 (1996)). LH
R was expressed by follicular membrane cells, granulosa cells of preovulatory follicles and luteinizing granulosa cells of the corpus luteum (Sirois, J., et al., J. Biol. Chem., 267: 11586-11592 (1992). )).
In GDF-9 deficient ovaries, non-luteinized and lutealized follicular ovarian granulosa cells expressed LHR at very high levels. Stimulation of follicular and luteinizing granulosa cells by LH stimulates the production of the steroidogenic enzyme P-450scc, followed by the synthesis of progesterone (Richards, JS, Endocrine Reviews, 15: 725-751 (1994).
)). In wild-type ovaries, P-450scc protein was present in follicular membrane cells, corpus luteum and secondary stromal tissue. In GDF-9 deficient ovaries, P-450scc protein was detected at low levels in non-luteinized follicles and at much higher levels in steroidogenic "luteinized" follicles. GDF-9 female 6 weeks old
Deficient mice had an average serum progesterone level of 3.4 ng / ml, and compared to 2.6 ng / ml in wild-type mice, these nests not only expressed the marker, but also were "miniature". Shows that it also has the ability to function like the corpus luteum. However, as noted above, these follicular nests also express inhibin alpha, a marker that is never observed at significant levels in the corpus luteum.

Discussion The absence of GDF-9 in mammalian oocytes has previously been shown to lead to infertility by blocking in follicular development. As summarized in Figure 6, defects at the molecular level are further characterized. GDF-9
mRNA (McGrath, SA, et al., Mol. Endocr., 9: 131-136 (1995)) and protein (Elvin, JA et al., Mol. Endocr., (Submitted) (1999)) are added to primordial follicles. But is first observed in type 3a primary follicles. Ovaries from GDF-9 knockout mice show a block at the type 3b primary follicle stage. As shown herein, granulosa cells of type 3b follicles are essentially dormant: no cell division or apoptosis is observed in follicular granulosa cells until the oocytes disappear.
Thus, while the GDF-9 protein is synthesized at the type 3a stage, the GDF-9 signaling cascade should be essential only at the type 3b stage for further follicle maturation. Clearly, these experiments demonstrate the recruitment of primordial follicles and the development of granulosa cells from the primordial follicular stage (<20 granulosa cells) to the type 3b stage (90 granulosa cells) (Pedersen, T., "Follicle Growth"). in the Mouse Ovary ", In: SABigger JD (e
d) Oogenesis, University Park Press, Baltimore, pp. 361-376 (1972)) is GDF
Indicates that it does not depend on -9.

[0096] Unlike the dramatic apoptosis observed in regressive follicles of wild-type ovaries, G
Minimal apoptosis is observed in DF-9 knockout ovaries. Apoptosis usually occurs in the follicular cavernous follicles at a point after it becomes responsive and dependent on FSH. Even if GDF-9 deficient ovarian granulosa cells express the cavity marker, few apoptotic cells were found. These observations lead to some alternative explanations. GDF-9 is pro-apoptotic.
totic) may be required to elicit the ability to respond to a stimulus. Also, elevated serum FSH (Dong, J., et al., Nature, 383: 531-535 (1996)) cannot overcome type 3b block but can promote the survival of granulosa cells. Eventually, granulosa cells in the GDF-9 deficient ovaries may circumvent this "capable of apoptotic" state by differentiating after the oocytes have degenerated to form steroid-producing follicles. Thus, GDF-9 is an important factor for "differentiation" of granulosa cells, and post-type 3b granulosa cells acquire specific properties such as the ability to undergo apoptosis. Becomes possible.

It has previously been hypothesized that oocyte-derived factors regulate kit ligand expression by a paracrine mechanism (Packer, AI et al., Dev. Biol., 161: 194-205).
(1994)). In mice stimulated with gonadotropin, there is a gradient of kit ligand expression, whereby granulosa cells farthest from the oocyte (ie, mural granulosa cells) are expressed at the highest level, while oocytes have the highest level. Near granulosa cells (ie, cumulus cells) express at very low or undetectable levels (Motro, B. and
Bernstein, A., Dev. Dyn., 197: 69-79 (1993)). GDF-
A dramatically elevated kit ligand in the 9-deficient follicle indicates that GDF-9 is one of the oocyte-secreted paracrine factors that negatively regulates kit ligand expression. This hypothesis is supported by evidence from our in vivo experiments of GDF-9 action presented herein, in which GDF-9 is another gene that is specifically expressed on follicular ovarian follicle oocytes. (I.e., hyaluronan synthase 2, cyclooxygenase 2, steroidogenic early regulatory protein, urokinase plasminogen activator and luteinizing hormone receptor). Thus, the effects of other oocyte-producing and extra-follicular factors unopposed by GDF-9 appear to contribute to the increased kit ligand expression observed here. Kit ligands produced in GDF-9 deficient ovaries also appear to be active. As previously reported in other systems, GDF-9 deficient ovaries contain significantly more mast cells per section, presumably due to kit ligands that stimulated recruitment and increased proliferation. Kit ligands have also been shown to stimulate oocyte development in vitro (Packer, AI et a
l., Dev. Biol., 161: 194-205 (1994)). Oocytes in GDF-9 knockout ovaries are more rapid and have a maximum size of 1 compared to controls before final degeneration.
Matured to 5% larger (Carabatsos, M., et al., Dev. Biol., 203: 373-384 (
1998)), provides further evidence of functional granulosa cell-inducing kit ligand signaling through c-kit on oocytes. As shown herein by Northern blot analysis, other members of the TGF-β superfamily continue to be expressed in GDF-9 deficient ovaries. By Northern blot analysis,
The inhibin α and activin βA subunits are expressed at levels similar to controls in GDF-9 deficient ovaries, but activin βB and follistatin are dramatically reduced. Activin A stimulates follicle maturation in vitro (Mather e
t al., 1997) and enabling FSH stimulation of granulosa cell DNA synthesis (Miro
, F. and Hillier, S., Endocr., 137: 464-468 (1996)). Furthermore, activin A + FSH stimulates granulosa cells from immature follicles to produce progesterone, but progesterone synthesis was reduced by granulosa cells from differentiated cultures with or without FSH (Miro, F., et al., Endocr., 129
: 3388-3394 (1991)). In GDF-9 deficient ovaries, activin A produced locally in the follicle nest stimulates the proliferation of restricted granulosa cells and enhances the response of these cells to elevated serum gonadotropin to increase P450 aromatase and P450 aromatase levels. -4
It appears to express 50 scc. Activin effects in GDF-9 deficient ovaries also appear to be enhanced by lower levels of activin binding protein and the antagonist follistatin.

Example 5 Identification of the Stimulatory Effect of Recombinant mGDF-9 on Progesterone Synthesis by Mural Granulosa Cells Cultured In Vitro The effect of mGDF-9 on progesterone production was determined by follicle stimulating hormone (FS
It occurs only in the absence of H) or at low doses (0.5 or 1 ng / ml). At higher doses (5 ng / ml and 10 ng / ml) of FSH, only FSH or FSH
There is no significant difference in progesterone production with + mGDF-9. FSH appears to stimulate P450 cholesterol side-chain cleavage expression, an enzyme that catalyzes a rate-limiting step in the cholesterol biosynthesis pathway, as previously shown (Oonk, et al.). Based on the most recent data, progesterone stimulation by GDF-9, not P450scc, is a steroidogenic early regulatory protein, an enzyme that has been shown to be involved in mobilizing or transporting cholesterol into cells. (STAR) appears to be due to stimulation.

Example 6 Identification of the Stimulatory Effect of Recombinant mGDF-9 on Follistatin and Activin βB Expression by In Vitro Cultured Mural Granulosa Cells Protocol Primary granulosa cells were isolated from the eight pairs of ovaries described above. 50-60 ng / ml in an α-MEM based medium containing 4 ng / ml FSH and no fetal calf serum
Cultured for 5 hours with or without recombinant mGDF-9. After the culture, remove the medium,
Non-adherent cells were collected by centrifugation. RNA was isolated using the Qiagen RNEasy total RNA isolation kit. This RNA was reverse transcribed into cDNA used to generate probes for hybridization to the Affymetrix developmental biology 250 gene chip. Based on the results of this initial screening method, the granulosa cell culture experiment was repeated, and RNA was prepared and used for Northern blot analysis.

Results According to developmental chip analysis, recombinant mGDF-9 showed a 4.9-fold increase in follistatin expression by primary granulosa cells cultured in vitro Δ191 (mGDF-9
Treatment): 39 (control)｝ and 3.3-fold increase in activin βB expression ｛65 (
mGDF-9 treatment): 20 (control) were stimulated (Fig. 3 / 15-1)
). This result was confirmed by Northern blot analysis (Fig. 3 /
15-2). Densitometer analysis of these Northern blots revealed mG
DF-9 treatment increased follistin expression 3.1 fold and activin βB expression 3.
Stimulation to a 3-fold increase was indicated. Furthermore, follistatin expression was determined by mGDF-
9 in a dose-dependent manner (FIGURE 3 / 15-3). These results indicate that in vivo in the absence of GDF-9, a three-fold reduction in follistin expression and an approximately six-fold reduction in activin βB expression indicate total RNA isolated from ovaries of our GDF-9 deficient mice. This is further supported by Northern blot analysis.

Significance The current data showing the regulation of follistatin and activin βB by mGDF-9 suggests our earlier results that the recombinant mGDF-9 protein we are making is biologically active. Is confirmed. Radioimmunoassays are available to analyze protein production for both follistatin and activin βB and provide two additional assays for GDF-9 agonists and antagonists. Furthermore, these results provide two additional downstream targets that can be used to identify GDF-9 response elements for transcriptional regulation and to understand GDF-9 signaling.

Example 7 Protocol for Identifying the Inhibitory Effect of Recombinant mGDF-9 on Luteinizing Hormone Receptor (LHR) Expression by Mural Granulosa Cells In Vitro Cultured Granulosa cells were isolated and described as above. Fetal bovine serum for various periods
/ -FSH. RNA was isolated as described and RT-PCR was performed on the LHR message as described using primers that span the intron.

Results LHR expression increases with increasing culture time in control medium (Figu
re 3 / 15-4). However, LHR expression is suppressed in media treated with 50 ng / ml mGDF-9.

Significance LHR expression in vivo is rare, if detected, on granulosa cumulus cells (in direct contact with the oocyte) and increases distance from the oocyte Has previously been shown to increase. Recent studies (Eppig, et al., 1997) have shown that oocyte secreted factors have the ability to inhibit LHR expression. The results described herein indicate that GDF-9 is an oocyte secreted factor that normally inhibits LHR expression.

Equivalents The present invention has been particularly shown and described with respect to preferred embodiments thereof, but without departing from the spirit and scope of the invention, which is defined by the appended claims. It will be understood by those skilled in the art that various changes can be made in form and detail. Those skilled in the art will recognize, using no more than routine experimentation, many equivalents to the specific embodiments of the invention described in detail herein.
And will be able to confirm. Such equivalents are intended to be encompassed by the following claims.

[Brief description of the drawings]

FIG. 1. FIG. 1 shows that conditioned media containing recombinant mouse GDF-9 (mGDF-9) or recombinant human GDF-9 (hGDF-9) was cultured in vitro in a time and concentration dependent manner. FIG. 5 is a bar graph of the results of a progesterone radioimmunoassay showing stimulation of progesterone synthesis by primary granulosa cells.

FIG. 2 is a bar graph of the results of a progesterone radioimmunoassay showing that recombinant mGDF-9 stimulated progesterone synthesis by primary granulosa cells in vitro.

FIG. 3 shows recombinant mGDF with and without fetal calf serum (FCS).
9 is a bar graph of the results of a progesterone radioimmunoassay showing that -9 stimulated progesterone synthesis by primary granulosa cells in vitro for 24 hours.

FIG. 4 shows various concentrations of FSH at 0, 0.5, or 1 ng / ml.
Is a bar graph showing the level of progesterone in the media of granulosa cells cultured in serum-free media for 14 hours in the presence or absence of 50 ng / ml GDF-9 in the presence of GDF-9; error bars are the mean (SEM). Shows the standard error. ^* , Control vs. FSH = 0 ng / ml, FSH = 0.5 ng / ml, or FSH = 1 ng / m
p <0.05 for 50 ng / ml GDF-9 at 1.

FIG. 5 shows the level of progesterone in the medium after treatment with cells +/− GDF-9 (100 ng / ml) or FSH for 24 hours for three separate samples of medium containing 10% FCS. It is a bar graph shown. ^* , P <0.05 for control versus GDF-9 treated cells cultured without FSH.

FIG. 6 shows Figures 4C to 4D that are normalized to HPRT levels.
4A is a bar graph showing quantification of P-450scc and StAR mRNA levels from 4A and 4B experiments; each value is the mean of two or three separate wells +/−.
SEM; FSH = 0 or 10 ng / m in StAR mRNA analysis.
Control treated sample at 1 (n = 4).

FIG. 7 is a model showing the role of GDF-9 in mammalian ovaries.

FIG. 8 is a summary of GDF-9 knockout ovary studies.

──────────────────────────────────────────────────続 き Continuation of front page (81) Designated country EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE ), OA (BF, BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, GM, KE, LS, MW, SD, SL, SZ, UG, ZW), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AE, AL, AM, AT, AU, AZ, BA, BB, BG, BR , BY, CA, CH, CN, CU, CZ, DE, DK, EE, ES, FI, GB, GD, GE, GH, GM, HR, HU, ID, IL, IN, IS , JP, KE, KG, KP, KR, KZ, LC, LK, LR, LS, LT, LU, LV, MD, MG, MK, MN, MW, MX, NO, NZ, PL, PT, RO, RU, SD, SE, SG, SI, SK, SL, TJ, TM, TR, TT, UA, UG, US, UZ, VN, YU, ZA, ZW (72) Inventor Elvin, Julia, A. United States Texas 77025 Houston, Stanton Street 3114 (72) Inventor One, Pay United States Texas 77025 Houston, Number Nine, Glen Arbor 3822 F-term (reference) 2G045 AA40 BB20 CB01 CB26 CB30 DA20 DA30 DA36 DA54 DA59 DA69 DA80 FA14 FB02 FB03 FB04 FB05 FB06 FB07 FB08 FB11 FB15

Claims (46)

[Claims] 1. a) i) GDF-9 receptor and GDF to the receptor
A cell having a gene whose expression is regulated by the binding of -9; ii) GDF-9; and iii) a drug to be evaluated; b) conditions suitable for binding of GDF-9 to a receptor on the cell. Maintaining the combination prepared in (a); and c) measuring the degree of expression of the gene regulated by binding of GDF-9 to the receptor that occurs upon binding of GDF-9 to the receptor. A change in the expression of the gene upon binding of GDF-9 to the receptor in the presence of the drug to be evaluated, indicating that the drug alters GDF-9 activity. -9) A method for identifying an agent that alters the activity. 2. The method of claim 1, wherein the extent to which expression of the gene occurs is determined by measuring the gene product. 3. The method according to claim 1, wherein the gene is hyaluronan synthase, a steroidogenic early regulatory protein.
y protein), luteinizing hormone receptor, cyclooxygenase 2
3. The method of claim 2, which encodes a protein selected from the group consisting of urokinase plasminogen activator, kit ligand, activin / inhibin βB and follistatin. 4. The method according to claim 1, wherein the extent to which the expression of the gene occurs is determined by measuring the product or function resulting from the activity of the gene in a) i). 5. The product or function as a result of binding of hyaluronic acid, a phosphorylated steroid production early regulatory protein, progesterone, plasmin, prostaglandin, activin βB: activin βB, inhibin α: activin βB, luteinizing hormone receptor. 5. The method of claim 4, wherein the product is selected from the group consisting of a product produced, a product produced as a result of binding of a ligand to a luteinizing hormone receptor, and a product produced as a result of cleavage of plasminogen. 6. The method of claim 1, further comprising the step (d) of comparing the degree of expression in the control sample with the degree of expression in (c). 7. a) The cells of i) are mural granulosa cells (mural gra
The method according to claim 1, wherein the method is selected from the group consisting of nulosa cells and cumulus cells. 8. a) combining i) granulosa cells; ii) GDF-9; and iii) the agent to be evaluated; b) conditions suitable for binding GDF-9 to a receptor on the granulosa cells. Maintaining the combination prepared in (a); and c) expression of a gene regulated by binding of GDF-9 to the receptor resulting upon binding of GDF-9 to the receptor on granulosa cells Measuring the degree of the GDF-9, wherein a change in expression of the gene upon binding of GDF-9 to the receptor in the presence of the drug to be evaluated indicates that the drug alters GDF-9 activity.
A method for identifying an agent that alters the activity of -9. 9. The method of claim 8, wherein the degree to which expression of the gene occurs is determined by measuring the gene product. 10. A protein whose gene is selected from the group consisting of hyaluronan synthase, steroidogenic early regulatory protein, luteinizing hormone receptor, cyclooxygenase 2, urokinase plasminogen activator, kit ligand, activin / inhibin βB and follistatin. 10. The method of claim 9, wherein 11. The method of claim 8, wherein the extent to which the expression of the gene occurs is determined by measuring the product or function resulting from the activity of the gene in a) i). 12. The product or function as a result of binding of hyaluronic acid, a phosphorylated steroid production early regulatory protein, progesterone, plasmin, prostaglandin, activin βB: activin βB, inhibin α: activin βB, luteinizing hormone receptor. 12. The method of claim 11, wherein the product is selected from the group consisting of a product produced, a product produced as a result of binding of a ligand to a luteinizing hormone receptor, and a product produced as a result of cleavage of plasminogen. 13. The degree of expression occurring in the control sample and (c)
9. The method according to claim 8, comprising the step (d) of comparing the degree of expression in 14. a) combining i) granulosa cells; ii) GDF-9; and iii) the agent to be evaluated; b) conditions suitable for binding GDF-9 to a receptor on the granulosa cells. Maintaining the combination prepared in (a); and c) expression of a gene regulated by binding of GDF-9 to the receptor resulting upon binding of GDF-9 to the receptor on granulosa cells Measuring the extent of, wherein inhibiting the expression of the gene upon binding of GDF-9 to the receptor in the presence of the agent to be evaluated indicates that the agent inhibits GDF-9 activity, GD
A method for identifying a drug that is an inhibitor of F-9 activity. 15. The method of claim 14, wherein the extent to which expression of the gene occurs is determined by measuring the gene product. 16. The method of claim 15, wherein the gene encodes a protein selected from the group consisting of hyaluronan synthase, steroidogenic early regulatory protein, luteinizing hormone receptor, cyclooxygenase 2, activin / inhibin βB, and follistatin. . 17. The method of claim 14, wherein the extent to which expression of the gene occurs is determined by measuring the product or function resulting from the activity of the gene in a) i). 18. The method according to claim 18, wherein the product or function is hyaluronic acid, phosphorylated steroid production early regulatory protein, progesterone, prostaglandin, activin βB.
18. Activin βB, Inhibin α: Activin βB, selected from the group consisting of products produced as a result of binding of luteinizing hormone receptor and products produced as a result of binding of ligand to luteinizing hormone receptor.
The described method. 19. The degree to which expression occurs in the control sample and (c)
The method according to claim 14, comprising a step (d) of comparing the degree of expression in 20) combining a) i) granulosa cells; ii) GDF-9; and iii) the agent to be evaluated; b) conditions suitable for binding GDF-9 to a receptor on the granulosa cells. Maintaining the combination prepared in (a); and c) the regulation of the uPA gene regulated by binding of GDF-9 to the receptor resulting upon binding of GDF-9 to the receptor on granulosa cells. Measuring the degree of expression, wherein increasing the production of uPA in the presence of
A method for identifying a drug that is an inhibitor of GDF-9 activity, wherein the agent is an inhibitor of -9 activity. 21) combining a) i) granulosa cells; ii) GDF-9; and iii) the agent to be evaluated; b) conditions suitable for binding GDF-9 to a receptor on the granulosa cells. Maintaining the combination prepared in (a); c) expression of a gene regulated by binding of GDF-9 to the receptor resulting upon binding of GDF-9 to the receptor on granulosa cells Measuring the extent, wherein an increase in the expression of the gene upon binding of GDF-9 to the receptor in the presence of the drug to be evaluated indicates that the drug is an enhancer of GDF-9 activity, A method for identifying a drug that is an enhancer GDF-9 activity. 22. The method of claim 21, wherein the extent to which expression of the gene occurs is determined by measuring the gene product. 23. The method of claim 22, wherein the gene encodes a protein selected from the group consisting of hyaluronan synthase, a steroidogenic early regulatory protein, luteinizing hormone receptor, cyclooxygenase 2, activin / inhibin βB, and follistatin. . 24. The method of claim 21, wherein the extent to which expression of the gene occurs is determined by measuring the product or function resulting from the activity of the gene in a) i). 25. The product or function of which is hyaluronic acid, phosphorylated steroid production early regulatory protein, progesterone, prostaglandin, activin βB
25. The method of claim 24, wherein the method is selected from the group consisting of: activin βB, a product produced as a result of binding of a luteinizing hormone receptor, and a product produced as a result of binding of a ligand to a luteinizing hormone receptor. 26. The degree of expression occurring in the control sample and (c)
22. The method according to claim 21, comprising the step (d) of comparing the degree of expression in 27) combining a) i) granulosa cells; ii) GDF-9; and iii) the agent to be evaluated; b) conditions suitable for binding GDF-9 to a receptor on the granulosa cells. Maintaining the combination prepared in (a); and c) binding of the GDF-9 to the receptor on the granulosa cells resulting in the regulation of the uPA gene by the binding of GDF-9 to the receptor. Measuring the degree of expression, wherein the reduction of uPA production in the presence of
A method for identifying a drug that is an enhancer of GDF-9 activity, the method being an enhancer of GDF-9 activity. 28) a step of combining a) i) granulosa cells; ii) GDF-9; and iii) an agent to be evaluated; b) conditions suitable for binding GDF-9 to a receptor on the granulosa cells Maintaining the combination prepared in (a); and c) expression of a gene regulated by binding of GDF-9 to the receptor resulting upon binding of GDF-9 to the receptor on granulosa cells Measuring the extent of the fertility, wherein inhibiting expression of the gene upon binding of GDF-9 to the receptor in the presence of the agent to be evaluated indicates that the agent inhibits fertility. Method for identifying a drug that inhibits ATP. 29. The method of claim 28, wherein the extent to which expression of the gene occurs is determined by measuring the gene product. 30. The method of claim 29, wherein the gene encodes a protein selected from the group consisting of hyaluronan synthase, steroidogenic early regulatory protein, luteinizing hormone receptor, cyclooxygenase 2, activin / inhibin βB, and follistatin. . 31. The method of claim 28, wherein the extent to which expression of the gene occurs is determined by measuring the product or function resulting from the activity of the gene in a) i). 32. The product or function of which is hyaluronic acid, phosphorylated steroidogenesis early regulatory protein, progesterone, prostaglandin, activin βB
32. Activin βB, inhibin α: Activin βB, a product produced as a result of binding of a luteinizing hormone receptor and a product produced as a result of binding of a ligand to a luteinizing hormone receptor.
The described method. 33. The degree of expression occurring in the control sample and (c)
29. The method according to claim 28, comprising the step of comparing the degree of expression in (d). 34) combining i) a granulosa cell; ii) GDF-9; and iii) an agent to be evaluated; b) conditions suitable for binding GDF-9 to a receptor on the granulosa cell. Maintaining the combination prepared in (a); and c) the uPA gene regulated upon binding of GDF-9 to the receptor, which occurs upon binding of GDF-9 to the receptor on granulosa cells A method of identifying a drug that inhibits fertility, wherein an increase in uPA production in the presence of a high concentration of GDF-9 indicates that the drug inhibits fertility. . 35) combining a) i) granulosa cells; ii) GDF-9; iii) a drug to be evaluated; b) under conditions suitable for binding of GDF-9 to a receptor on the granulosa cells. (A) maintaining the combination prepared in (a); and c) expression of a gene regulated by binding of GDF-9 to the receptor that occurs upon binding of GDF-9 to the receptor on granulosa cells. Measuring the extent, wherein enhancing expression of the gene upon binding of GDF-9 to the receptor in the presence of the agent to be evaluated indicates that the agent enhances fertility, Methods for identifying potentiating agents. 36. The method of claim 35, wherein the degree to which expression of the gene occurs is determined by measuring the gene product. 37. The method of claim 36, wherein the gene encodes a protein selected from the group consisting of hyaluronan synthase, steroidogenic early regulatory protein, luteinizing hormone receptor, cyclooxygenase 2, activin / inhibin βB, and follistatin. . 38. The method of claim 35, wherein the extent to which expression of the gene occurs is determined by measuring the product or function resulting from the activity of the gene in a) i). 39. The product or function of the protein is hyaluronic acid, phosphorylated steroid production early regulatory protein, progesterone, prostaglandin, activin βB
39. An activin βB, inhibin α: activin βB, selected from the group consisting of a product produced as a result of binding of a luteinizing hormone receptor and a product produced as a result of binding of a ligand to a luteinizing hormone receptor.
The described method. 40. The degree of expression occurring in the control sample and (c)
36. The method of claim 35, comprising the step of comparing the degree of expression in (d). 41) combining a) i) granulosa cells; ii) GDF-9; and iii) the agent to be evaluated; b) conditions suitable for binding GDF-9 to a receptor on the granulosa cells. Maintaining the combination prepared in (a); and c) the regulation of the uPA gene regulated by binding of GDF-9 to the receptor resulting upon binding of GDF-9 to the receptor on granulosa cells. Measuring the degree of expression, wherein a reduction in uPA production in the presence of high concentrations of GDF-9 indicates that the agent enhances fertility. 42) a) i) GDF-9 receptor and GD for the receptor
Combining a cell having a gene whose gene expression is regulated by F-9 binding; and ii) an agent to be evaluated; b) under conditions suitable for binding GDF-9 to a receptor on the cell (a C) determining the extent to which expression of the gene regulated by the binding of GDF-9 to the receptor occurs, and c) determining the degree of expression of the gene in the presence of the drug to be evaluated. Expression of GDF-9
A method for identifying a drug that is an agonist of GDF-9, wherein the drug is an agonist of GDF-9. 43. The method of claim 42, wherein the agonist enhances fertility. 44) a) i) GDF-9 receptor and GD for the receptor
A cell having a gene whose expression is regulated by F-9 binding; ii) GDF-9; and iii) a drug to be evaluated; b) conditions suitable for binding the drug to a receptor on the cell Maintaining the combination prepared in a) below; and c) determining the extent to which binding of the drug to the receptor on the cell occurs, wherein the binding of the drug to the GDF-9 receptor is A method for identifying a drug that is an agonist of GDF-9, wherein the method indicates that the drug is an agonist of GDF-9. 45) a) i) GDF-9 receptor and GD for said receptor
A cell having a gene whose expression is regulated by F-9 binding; ii) GDF-9; and iii) a drug to be evaluated; b) conditions suitable for binding the drug to a receptor on the cell Maintaining the combination prepared in a) below; and c) determining the extent to which binding of the drug to the receptor on the cell occurs, wherein the binding of the drug to the GDF-9 receptor is A method for identifying a drug that is an antagonist of GDF-9, wherein the method indicates that the drug is an antagonist of GDF-9. 46. The method of claim 46, wherein said antagonist inhibits fertility.