IGF1R Expression in Ovarian Granulosa Cells Is Essential for Steroidogenesis, Follicle Survival, and Fertility in Female Mice

Abstract

Folliculogenesis is a lengthy process that requires the proliferation and differentiation of granulosa cells (GCs) for preovulatory follicle formation. The most crucial endocrine factor involved in this process is follicle-stimulating hormone (FSH). Interestingly, previous in vitro studies indicated that FSH does not stimulate GC proliferation in the absence of the insulinlike growth factor 1 receptor (IGF1R). To determine the role of the IGF1R in vivo, female mice with a conditional knockdown of the IGF1R in the GCs were produced and had undetectable levels of IGF1R mRNA and protein in the GCs. These animals were sterile, and their ovaries were smaller than those of control animals and contained no antral follicles even after gonadotropin stimulation. The lack of antral follicles correlated with a 90% decrease in serum estradiol levels. In addition, under a superovulation protocol no oocytes were found in the oviducts of these animals. Accordingly, the GCs of the mutant females expressed significantly lower levels of preovulatory markers including aromatase, luteinizing hormone receptor, and inhibin α. In contrast, no alterations in FSH receptor expression were observed in GCs lacking IGF1R. Immunohistochemistry studies demonstrated that ovaries lacking IGF1R had higher levels of apoptosis in follicles from the primary to the large secondary stages. Finally, molecular studies determined that protein kinase B activation was significantly impaired in mutant females when compared with controls. These in vivo findings demonstrate that IGF1R has a crucial role in GC function and, consequently, in female fertility.

Infertility is a common medical problem affecting >10% of reproductive-aged women, of whom ~40% have ovulatory dysfunction (1, 2). Ovulation results from the lengthy folliculogenesis process that includes primordial follicle recruitment, followed by granulosa cell (GC) proliferation and differentiation, leading to antral or preovulatory follicle formation. Unequivocally, the most crucial endocrine factor involved in the stimulation of follicle growth and survival is follicle-stimulating hormone (FSH), which targets GCs exclusively to induce their maturation and differentiation (3). Among the most important local factors is the insulinlike growth factor (IGF) system. Our previous in vitro studies in mouse, rat, and human GCs demonstrated that the stimulatory effect of FSH on aromatase expression, a marker of GC differentiation, depends on IGF1 (rodents) or IGF2 (humans) action and the expression and activation of the IGF1 receptor (IGF1R) (4–6). These findings have been confirmed by studies using undifferentiated rat GCs, in which pharmacological inhibition of IGF1R activity prevents FSH phosphorylation of protein kinase B (Akt) (7). Also, in bovine GCs, FSH acts synergistically with IGF1 to increase cell number and aromatase expression (8). In addition, it is known that IGF1 enhances the stimulatory effects of FSH on progesterone production (9), aromatase/estradiol (10), luteinizing hormone receptor (11), and inhibin-α (12). The synergistic effects of IGF1 and FSH on GC differentiation indicate that these hormones cooperate to maintain GC function. Therefore, understanding the interaction between FSH and the IGF system may lead to advances in infertility treatments by exploiting the mechanisms that coordinate these signals.

Animal studies have shown that FSH and IGFs are crucial for the development of preovulatory follicles. FSH- and FSH receptor–deficient female mice are infertile because of an arrest of folliculogenesis at the secondary stage (3, 13). Similarly, follicle growth stops at the antral stage in the few IGF1 knockout mice that survive to adulthood (14, 15). However, IGF1 knockout mice exhibit growth deficiency, and depending on the genetic background, only 10% to 60% survive to adulthood (16). This evidence suggests that infertility in IGF1 knockout mice could be caused by multiple factors that contribute to a decrease in the well-being of the animals. On the other hand, IGF1R knockout mice die at birth (16); therefore, the role of the IGF1R in ovarian function and female fertility remains to be determined.

The aim of this investigation was to determine the consequences of the conditional deletion of the IGF1R in ovarian GCs on folliculogenesis and fertility. Specifically, we tested the hypothesis that knockdown of IGF1R expression in GCs results in loss of responsiveness to FSH in preantral follicles. Our findings demonstrate a crucial role of the IGF1R in vivo in the regulation of follicle growth and the survival and differentiation of GCs. These findings significantly contribute to our understanding of the mechanisms by which FSH and IGFs interact to control GC function.

Materials and Methods

Animals

Animals were treated in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and all protocols were approved by the University of Illinois at Chicago Animal Care Committee. Mice of the following strains were used: IGF1R-F/F, INSRF/F, and ZP3-Cre mice were obtained from the Jackson Laboratory (Bar Harbor, ME). Cyp19-Cre and Esr2-Cre mice were previously described (17–19).

GC isolation

Immature female mice between 21 and 25 days old were injected intraperitoneally with 7.5 IU pregnant mare serum gonadotropin [equine chorionic gonadotropin (eCG)] (Sigma-Aldrich, St. Louis, MO), and the ovaries were collected after 48 hours. For ovary collection, mice were anesthetized with isoflurane, followed by cervical dislocation before ovary removal. Ovaries were dissected and placed in M199 media containing 6.8 mM ethylene glycol tetraacetic acid solution for 10 minutes. Then ovaries were transferred to M199 media containing 0.5 M sucrose, again for 10 minutes. Follicles were then punctured to extrude granulosa cells, which were then strained and spun down at room temperature for 10 minutes at 1000g. Cells were resuspended in either TRIzol reagent for RNA isolation or radioimmunoprecipitation assay buffer for protein isolation.

RNA isolation and quantitative real-time polymerase chain reaction

Total RNA was isolated from primary granulosa cells with TRIzol Reagent (Invitrogen, Carlsbad, CA) used as specified in the manufacturer’s protocol. Total RNA (1 µg) was reverse transcribed via Moloney murine leukemia virus reverse transcription (Invitrogen) and anchored oligo-dT primers (Integrated DNA Technologies, Inc., Coralville, IA) at 42°C for 1 hour. The resulting complementary DNA was diluted to a final concentration of 10 ng/µL, and quantitative real-time polymerase chain reaction (PCR) was used to measure the expression Rpl19 and transcripts of interest as described previously (18, 19).

Hematoxylin and eosin staining

Immature female mice between 21 and 25 days old were injected subcutaneously with 5 IU pregnant mare serum gonadotropin (eCG) (Sigma-Aldrich), and the ovaries were collected after 48 hours. Ovaries were fixed in formalin before paraffin embedding. Dewaxed and rehydrated tissue sections were stained with Harris Hematoxylin (Thermo Fisher Scientific, Waltham, MA) and counterstained with Eosin-Y and Phloxine (Thermo Fisher Scientific).

Immunohistochemistry

Ovaries were fixed in Bouin’s solution before paraffin embedding. Five-micron sections were dewaxed and rehydrated. This step was followed by antigen retrieval via citrate buffer solution (10 mM of citric acid and sodium citrate, pH 6) microwaved on high for 30 seconds until boiling and then at low for 8 more minutes. After cooling, slides were placed in 1% H₂O₂. Sections were then blocked with the Avidin/Biotin Blocking Kit (Vector Laboratories, Burlingame, CA) followed by 30 minutes of blocking in SuperBlock Blocking Buffer (Pierce Chemicals, Rockford, IL) before the addition of the primary antibody diluted in phosphate-buffered saline. After washes with Tris phosphate-buffered saline, slides were incubated with secondary antibody for 30 minutes at room temperature followed by washing. Tissues were stained with an Vectastain Elite ABC Kit or Red Alkaline Phosphatase (Vector Laboratories) according to manufacturer’s recommendations. Slides were counterstained with Gill’s hematoxylin before mounting.

Estradiol measurements

Truncal blood from control and experimental mice was centrifuged, and the serum fraction was collected for hormone measurements. Serum levels of estradiol were measured with estradiol enzyme-linked immunosorbent assay kits (DRG Instruments GmbH, Marburg, Germany) according to the manufacturer’s protocol. The range of this assay is between 9.7 and 2000 pg/mL. The intra-assay coefficient of variation is 4.55 ± 0.98 [mean ± standard error of the mean (SEM)], and interassay coefficient of variation is 7.8 ± 0.6. Undiluted serum was used.

Ovulation assay

Immature female mice between 21 and 25 days old were injected subcutaneously with 5 IU eCG (Sigma-Aldrich) followed by 5 IU human chorionic gonadotropin (hCG) (Sigma-Aldrich) 48 hours later. After 17 hours, mice were killed and the oviducts isolated. The cumulus–oocyte complexes were extruded from each oviduct into media containing hyaluronic acid to disperse cells from the oocytes. The denuded oocytes were then counted.

Western blotting

GCs were harvested in cold radioimmunoprecipitation assay lysis buffer supplemented with protease inhibitor cocktail (Sigma) and Halt phosphatase inhibitor cocktail (Thermo Fisher Scientific). Protein quantification and western blotting of whole cell lysates were carried out as previously described (Table 1) (19). The band intensities were quantified in Image Laboratory software (Bio-Rad Laboratories, Hercules, CA) and adjusted relative to β-actin.

Table 1.

Antibodies Used

Peptide/Protein Target	Antigen Sequence (if Known)	Name of Antibody	Antibody Research Resource Identifier	Manufacturer, Catalog #, or Name of Person Providing the Antibody	Species Raised in; Monoclonal or Polyclonal	Dilution Used
Total Akt	Synthetic peptide corresponding to the carboxy-terminal sequence of mouse Akt	Akt antibody	AB_329827	Cell Signaling, #9272	Rabbit; polyclonal	1:1000
Phospho-(S473)-Akt	Synthetic phosphopeptide corresponding to residues around Ser473 of human Akt	Phospho-Akt (Ser473) (D9E) XP® Rabbit mAbv	AB_2315049	Cell Signaling, #4060	Rabbit; monoclonal	1:2000
FSH receptor	Purify protein	Anti-FSH receptor antibody	AB_10899034	Abcam, #ab103874	Rabbit; polyclonal	1:5000
Cleaved caspase 3	Synthetic peptide corresponding to amino-terminal residues adjacent to (Asp175) in human caspase 3	Cleaved caspase 3 (Asp175) antibody	AB_2341188	Cell Signaling, #9661	Rabbit; polyclonal	1:200
Ki67	Recombinant protein specific to the amino terminus of human Ki-67 protein	Ki-67 (D2H10) rabbit mAb (IHC-specific)	AB_2636984	Cell Signaling, #9027	Rabbit; polyclonal	1:500
β-actin	Ac-Asp-Asp-Asp-Ile-Ala-Ala-Leu-Val-Ile-Asp-Asn-Gly-Ser-Gly-Lys	Antib-actin antibody	AB_476692	Sigma, A1978	Mouse; monoclonal	1:5000

Abbreviation: IHC, immunohistochemistry.

Statistics

Data are expressed as the mean ± SEM. Multiple group statistical analyses were performed by one-way analysis of variance followed by Tukey test for multiple comparisons. Two-group comparisons were performed via t test for independent samples. Statistics were calculated with GraphPad Prism 5 (GraphPad, La Jolla, CA).

Results

Conditional knockdown of the IGF1R in GCs

To understand the role of IGFs in the ovary without causing the severe defects reported in the mouse models that are globally deficient of IGF1 or IGF1R (16), we produced mice that lack IGF1R expression specifically in GCs. First, we produced mice carrying one floxed allele and one null allele for the IGF1R gene (denoted as IGF1R^f/−) and Cre recombinase under the control of either the estrogen receptor β (Esr2) promoter or the aromatase (Cyp19) promoter. In the ovary, Esr2 mRNA expression is confined to the GCs of preantral follicles (20, 21), whereas Cyp19Cre is active exclusively in the GCs of early antral follicles (22). Initial examination of these mice demonstrated that neither Esr2Cre nor Cyp19Cre is robust enough to decrease IGF1R mRNA levels in the GCs beyond 80% [Fig. 1(a)]. Considering that IGF1R expression as low as 10% of its normal levels is sufficient to maintain normal body functions including fertility (23), these findings suggest that the GCs of Esr2Cre-IGF1R^f/− and Cyp19Cre-IGF1R^f/− mice are still able to respond to IGF1.

Figure 1.

Knockdown of IGF1R expression in GCs. (a) IGF1R mRNA expression levels in GCs from eCG-treated animals expressed relative to mouse ribosomal L19 (the average of at least six samples per genotype is shown). Bars represent mean ± SEM (*P < 0.05). (b) Immunohistochemistry for the IGF1R protein in ovaries of 21- to 25-day-old controls and IGF1R^gcko eCG-stimulated mice (n = 3 for each genotype; a representative image is shown).

Because the combined expression of Esr2 and Cyp19 occurs only in GCs, we generated mice carrying Esr2Cre and Cyp19Cre to guarantee specific and maximal recombination of the IGF1R floxed allele. These Esr2⁺/Cyp19Cre-IGF1R^f/− mice will be referred to as IGF1R^gcko mice. Esr2⁺/Cyp19Cre-IGF1R^f/+ or Esr2⁺/Cyp19Cre-IGF1R^+/− mice were used as controls. In contrast to Esr2Cre-IGF1R^f/− or Cyp19Cre-IGF1R^f/− mice, IGF1R mRNA levels were undetectable by quantitative PCR in the GCs of IGF1R^gcko mice [Fig. 1(a)].

Because the ovarian expression and distribution of the IGF1R have not been reported previously, we performed immunohistochemistry for the IGF1R in the ovaries of control and IGF1R^gcko adult female mice. In control mice, IGF1R expression was detected in the GCs, the theca cells, the stroma, and the oocytes [Fig. 1(b), left]. The strongest staining for IGF1R was found in GCs of follicles from the primary to the antral stage. Higher magnifications confirmed the expected membrane localization of the receptor. Theca cells showed weak staining when compared with GCs, whereas the stroma displayed stronger staining than the theca cells but still significantly lower than GCs. Strong IGF1R expression was also present in the corpus luteum, but the signal was lower than in the follicles. In contrast, immunostaining for the IGF1R in the ovaries of IGF1R^gcko mice showed no signal in the GCs of all follicles, although IGF1R remained clearly detectable in the oocyte. IGF1R^gcko ovaries also showed weak staining in the stroma [Fig. 1(b), right, inset]. Together, these findings demonstrate that the knockdown of IGF1R is unique to the ovarian GCs in IGF1R^gcko mice.

Expression of the IGF1R in GCs is essential for reproduction

Fertility was tested in control and IGF1R^gcko females by mating them with males of proven fertility. To compare the effect of the different degrees of IGF1R knockdown observed in mice carrying Esr2Cre, Cyp19Cre, or their combination, we also examined the fertility of Cyp19Cre-IGF1R^f/− and Esr2⁺-IGF1R^f/− mice. The number of pups per litter and the total number of pups were recorded during 6 months of continued breeding. Control animals produced an average of 7 ± 0.1 pups per litter and 40 ± 3 pups over 6 months [Fig. 2(a)]. IGF1R^f/f and IGF1R^f/− mice had fertility profiles similar to those of control animals (data not shown). Cyp19Cre-IGF1R^f/f mice showed normal fertility compared with control animals (data not shown); however, Cyp19Cre-IGF1R^f/− animals were subfertile. For Cyp19Cre-IGF1R^f/−, the average litter size was 5.5 ± 0.5 pups, and the total number of pups per female over 6 months was 32 ± 2, both significantly lower when compared with control animals [Fig. 2(b)]. Two Esr2Cre⁺-IGF1R^f/− females produced several litters containing two to five pups, whereas two females produced no pups. These results indicate a high degree of variability in the fertility of female Esr2Cre⁺-IGF1R^f/− mice, although it is clear that these animals are subfertile. In sharp contrast, all IGF1R^gcko females were infertile and produced no pups over the entire breeding period. Taken together, these results indicate that IGF1R^gcko mice represent a unique platform to determine the role the IGF1R in GCs in vivo.

Figure 2.

IGF1R expression in GCs is essential for female fertility. (a) The number of pups per litter was determined in control mice and experimental animals over 6 months. Four or more females were used for each genotype. Columns represent the average ± SEM of the number of pups per litter (*P < 0.05; **P < 0.01 vs. controls). IGF1R^gcko females produced no pups. (b) Continuous breeding assessment is showing the cumulative number of progeny per female. Values represent the mean ± SEM of litters derived from at least five females for each genotype (*P < 0.05; **P < 0.01 vs. controls).

Next, we performed structural analyses in the ovaries of control, Cyp19Cre-IGF1R^f/−, and IGF1R^gcko mice after 48 hours treatment with eCG, which stimulates follicle growth via activation of FSH receptors. Follicles at all stages of development were present in control and Cyp19Cre-IGF1R^f/− ovaries, suggesting no gross defects in folliculogenesis. In contrast, no antral follicles were found in the ovaries of IGF1R^gcko mice after treatment with eCG [Fig. 3(a)].

Figure 3.

Abnormal folliculogenesis in IGF1R^gcko animals. (a) Representative hematoxylin and eosin staining of ovaries of control and IGF1R^gcko mice treated with eCG for 48 hours. (b) Estradiol levels in control (left) and experimental (right) mice treated with eCG for 48 hours. Bars represent the mean ± SEM of measurements from at least five animals per genotype. Different letters denote differences between genotypes (^{a–b, b–c}P < 0.05; ^a–cP < 0.01).

To examine the impact of the loss of the IGF1R on GC function, we measured serum estradiol levels in eCG-treated mice. Serum was collected from control, Cyp19Cre-IGF1R^f/f, Cyp19Cre-IGF1R^f/−, Esr2⁺/Cyp19Cre-IGF1R^f/f, and IGF1R^gcko mice 48 hours after treatment with eCG, and estradiol levels were measured by enzyme-linked immunosorbent assay. The results revealed no statistically significant differences between control, Cyp19Cre-IGF1R^f/f, and Cyp19Cre-IGF1R^f/− animals. Although estradiol levels were lower (∼40%) in Cyp19Cre-IGF1R^f/− animals, this difference did not reach statistical significance [Fig. 3(b), left]. In contrast, a significant reduction (>60%) in serum estradiol was observed in Esr2⁺/Cyp19Cre-IGF1R^f/f mice, which contain two floxed alleles of the IGF1R gene [Fig. 3(b), right]. A larger decrease in serum estradiol was observed in IGF1R^gcko mice, which carry a null and a floxed allele for the IGF1R. These results indicate that serum estradiol levels, and therefore follicular function, are directly proportional to the decline in IGF1R expression in GCs.

The lack of follicle development in IGF1R^gcko mice after treatment with eCG led us to hypothesize that one of the key mechanisms leading to infertility in these animals is an inability to release oocytes after the luteinizing hormone surge. To test this idea, we injected IGF1R^gcko mice with eCG followed 48 hours later by hCG administration, which mimics the cyclic luteinizing hormone surge causing synchronized ovulations. The oviducts were collected from control and experimental mice 17 hours after hCG, and the number of oocytes in the oviducts was determined. As expected, this protocol led to the release of many oocytes in control animals (Fig. 4). Cyp19Cre-IGF1R^f/− animals showed a tendency to ovulate fewer oocytes, but again the differences were not significant when compared with controls. In contrast, no oocytes were found in the oviducts of IGF1R^gcko mice.

Figure 4.

Lack of IGF1R expression in GCs blocks ovulation. Ovulation was induced in immature (22- to 23-day-old) animals by a subcutaneous injection of 5 IU eCG, followed 48 hours later by administration of 5 IU hCG. Oviducts and ovaries were harvested 17 hours after hCG; oocytes found in the oviduct at this time were counted. Bars represent mean ± SEM for five to seven animals per group. Data were compared via t test. No oocytes were found in the oviducts of IGF1R^gcko animals.

Insulin actions in GCs are not crucial for fertility

IGFs are known to bind the insulin receptor (INSR) (24). Accordingly, receptor-binding assays showed that IGF1 can reduce the binding of insulin to the INSR by 50% (25, 26). Therefore, it is possible that the activation of INSR could mediate some of the effects of IGFs in GCs. To determine the relative contribution of the INSR to GC function in vivo, we developed animals carrying floxed alleles for the INSR and both Cyp19Cre and Esr2-Cre⁺ (INSR^gcko). We performed fertility and superovulation studies by using the protocols described earlier. These studies showed that deletion of the INSR in the GCs has no major effect on the number of pups per litter or on the number of oocytes ovulated when compared with control animals (Table 2). These results suggest that insulin action in GCs is not essential for female fertility and that the IGF1R mediates IGF1 action in GCs.

Table 2.

Fertility of Female Mice Lacking INSR in the GCs

	Pups per Littera	Oocytes per Ovaryb
Control	6 ± 2 (n = 9)	10 ± 1 (n = 6)
INSRgcko	5 ± 1.3 (n = 4)	8 ± 2 (n = 3)

^a

Average number of pups per litter after a 6-month breeding period.

^b

Total number of oocytes released from each ovary after treatment with eCG and hCG.

The IGF1R is necessary for FSH-induced differentiation of GCs in vivo

The significant decrease in serum estradiol levels observed in IGF1R^gcko mice suggests that diminished IGF1R expression in GCs impairs FSH-induced steroidogenesis and differentiation. To confirm these findings and to further explore the role of the IGF1R, the expression of key steroidogenic genes and differentiation markers was quantified in the GCs of mice stimulated with eCG for 48 hours. The induction of steroidogenic genes by eCG treatment was significantly compromised in the GCs of IGF1R^gcko mice when compared with control animals [Fig. 5(a)]. Thus, the expression of StAR,Cyp11a1 (P450scc), and Cyp19a1 (aromatase) in the GCs of IGF1R^gcko mice was lower by sevenfold, 2.5-fold, and ninefold, respectively, when compared with control GCs.

Figure 5.

Relative expression of the main differentiation markers in GCs of control and IGF1R^gcko mice. (a) Total RNA was extracted from GCs isolated after eCG treatment in control and IGF1R^gcko females. The expression of genes important for GC differentiation was measured by quantitative PCR. Three or more animals were included for each genotype. Columns represent the mean ± SEM (***P < 0.01 vs. control). (b) FSH receptor (FSHR) mRNA (left) and protein levels (right). Blots are representative of three different animals. BACT, β-actin; StAR, steroidogenic acute regulatory protein.

The expression of markers of GC differentiation including luteinizing hormone receptor (LHR), inhibin-α, inhibin-Bα, and inhibin-Bβ were also significantly lower in GCs of IGF1R^gcko mice when compared with controls [Fig. 5(a)]. In contrast, the knockdown of the IGF1R had no effects on FSH receptor expression; thus, the GCs of control and IGF1R^gcko mice expressed comparable levels of FSH receptor protein or mRNA [Fig. 5(b)]. Together these findings demonstrate that IGF1R is critical for FSH-induced steroidogenesis and LHR induction.

Lack of IGF1R leads to apoptosis of GCs

The absence of antral follicles in the ovaries of IGF1R^gcko mice suggests that the proliferation and survival of the GCs may also be impaired. To examine this possibility, we costained control and IGF1R^gcko ovaries for markers of proliferation (Ki67, brown color) and apoptosis (cleaved caspase 3, pink color). Staining for Ki67 was observed in both control and IGF1R^gcko ovaries [Fig. 6(a)]. In contrast, increased staining for cleaved caspase 3 was found in IGF1R^gcko ovaries when compared with controls. Follicles at all stages of maturation, including secondary (arrowheads) and early antral (black arrow) follicles, showed extensive labeling (pink) for cleaved caspase 3. Staining for cleaved caspase 3 was also evident in follicles that showed extensive structural distortions (yellow arrows). Follicles with high degrees of structural regression were abundantly found in IGF1R^gcko ovaries but not in controls. These results suggest that GCs from IGF1R^gcko are prone to enter apoptosis, leading to a reduction in the number of GCs, which in turn diminishes follicle growth and promotes follicular atresia.

Figure 6.

Lack of IGF1R in GCs leads to increased apoptosis and diminished Akt activation. (a) Coimmunostaining for markers of proliferation (Ki67) and apoptosis (cleaved caspase 3) were performed in control and IGF1R^gcko ovaries from 23-day-old eCG-treated mice. Cleaved caspase 3 staining is depicted in pink, Ki67 staining is depicted in brown, and counterstaining by hematoxylin is depicted in light blue. Arrowheads indicate secondary follicles, black arrows indicate early antral follicles, and yellow arrows indicate follicles with structural distortion. High magnification (×20) sections are indicated (n = 3 for each genotype; a representative picture is shown). (b) Western blots for phospho-S473 and total Akt. *P < 0.001 vs. control, n = 3.

Defective in vivo activation of Akt by FSH in GCs lacking IGF1R

Our previous in vitro findings demonstrated that Akt is at the crossing of FSH and IGF1R signaling pathways (4, 5). Therefore, we next examined whether GCs of IGF1R^gcko mice have defective expression or activation of Akt. For this purpose, total Akt and phospho-S473-Akt were examined in the GCs of control and IGF1R^gcko mice treated with eCG for 1 hour. Akt phosphorylation was extremely low in the GCs of IGF1R^gcko mice compared with controls [Fig. 6(b)]; however, a noticeable decrease in total Akt was also observed. Analysis of these data revealed a significant reduction in the ratio of phosphorylated to total Akt in the GCs of IGF1R^gcko mice when compared with control cells [Fig. 6(b)], suggesting that the expression of the IGF1R in the GCs is crucial for the activation of Akt by FSH in vivo.

Discussion

Our findings demonstrate in vivo that GCs require the IGF1R to undergo differentiation upon FSH stimulation. Thus, FSH stimulation of steroidogenesis, estradiol production, and LHR expression depends on the presence of the IGF1R. The IGF1R is also essential for GC survival, and its absence results in increased apoptosis. These defects block folliculogenesis at the secondary stage, resulting in loss of fertility. This report also demonstrates that the IGFR is necessary for the formation of preovulatory follicles and the differentiation of GCs to the preovulatory stage. Thus, this in vivo evidence supports the presence of parallel and cooperative pathways between FSH and IGF1 in the control of growth, survival, and maturation of ovarian follicles.

A previous report suggested that the main role of the IGF system is to ensure a critical level of FSH receptor expression necessary for gonadotropin responsiveness, and the low FSH receptor expression is the reason for reduced follicle growth in IGF1 knockout mice (14). However, our findings indicate that expression of the FSH receptor remains unaffected in the GCs of IGF1R^gcko mice. This finding suggests that maintenance of FSH receptor levels is independent of the IGF system, in agreement with our previous in vitro findings that IGF1 has no effect on the expression of FSH receptor in rat GCs (4). Additionally, we have observed that inhibition of IGF1R activity blocks the stimulation of aromatase by agents that increase intracellular levels of cyclic adenosine monophosphate (cAMP) (4). Because cAMP is the main second messenger activated by FSH, this evidence demonstrates that the synergistic effects between FSH and the IGFs system on GC differentiation are mediated by signaling components downstream of cAMP. Supporting this idea, evidence in porcine GCs indicates that IGF1 stimulates cAMP response element binding protein activity (27). Moreover, in murine GCs, IGF1 amplifies cAMP-dependent increases in the mRNA abundance of important ovulatory response genes (28). However, the exact mechanisms involved in the effects of the IGF system on cAMP signaling remain to be investigated.

In addition to cAMP, FSH also stimulates the serine/threonine kinase Akt (29). Here, we show that FSH-induced Akt phosphorylation is minimal in the GCs of IGF1R^gcko mice. Akt is essential for the differentiation of GCs including the stimulation of 3β-hydroxysteroid dehydrogenase, α-inhibin, aromatase, LHR, cartilage link protein, and hypoxia-inducible factor 1 (30–32). Therefore, the failure of FSH to stimulate Akt phosphorylation in vivo is probably the most relevant defect that prevents the induction of GC differentiation in IGF1R^gcko mice. However, it is also known that activation of Akt alone in the absence of FSH is not enough to trigger GC differentiation (30). Therefore, the simultaneous activation of additional signaling molecules by the FSH receptor is necessary to trigger the differentiation program. More studies are needed to determine the molecular mechanisms downstream of Akt needed to sustain normal GC function and the integration of these mechanisms with other signaling pathways activated by the FSH receptor.

FSH is critical for the survival of antral follicles and plays an essential role in follicle selection where a limited number of follicles are selected for ovulation, whereas the remaining follicles undergo atresia. Thus, apoptosis and follicular atresia in hypophysectomized mice can be prevented by gonadotropin administration (33). However, our findings suggest that follicular apoptosis increases significantly in IGF1R^gcko mice even after treatment with gonadotropins. This indicates that the IGF1R is also necessary for the activation of the follicle survival program by FSH. This conclusion is in agreement with the finding that IGF binding protein 3, which reduces the availability of IGFs, partially reverses the suppression of apoptosis by FSH (33). Because Akt plays a crucial role in the regulation of cell survival (30), the increase in apoptosis observed in the GCs of IGF1R^gcko mice may result from the failure of these cells to activate Akt in response to FSH. Interestingly, in IGF1R^gcko mice, apoptosis was also evident in follicles at early stages of development, including primary and early secondary follicles, which are less reliant on FSH actions to survive. GC survival during early folliculogenesis could depend on the interaction of IGF and oocyte-secreted factors. More studies are needed to evaluate these hypotheses.

Insulin is regarded as a cogonadotropin in the ovary, mostly because of various in vitro studies documenting that it increases gonadotrophin-stimulated steroid synthesis in granulosa and theca cells (34–40). However, several of these studies use high concentrations of insulin, which can cross-react with the IGF1R or with IGF1R-INSR hybrids (41), suggesting that in these experiments the effect of insulin may be mediated by the activation of the IGF1R. Until now, few attempts have been made to determine the importance of insulin in ovarian function in vivo. Mice with reduced insulin receptor expression in the ovaries do not show significant alteration on folliculogenesis and fertility (42), suggesting that the presence of the INSR in GCs is not essential for female fertility. Our transgenic mice confirm these findings and demonstrate that knockout of insulin receptor specifically in GCs has no measurable effects on fertility. It is possible that the lack of INSR expression is compensated by signaling downstream of the IGF1R. If this is the case, our findings demonstrate that the absence of the IGF1R cannot be compensated by signaling downstream of the INSR.

In summary, our findings demonstrate a crucial role of the IGF1R in regulation of follicle growth and in survival and differentiation of GCs in vivo. The evidence indicates that one of the main consequences of the lack of IGF1R signaling in GCs is an increase in apoptosis at all stages of follicle development. In addition, follicles that develop to the secondary stage fail to respond to FSH. This prevents initiation of the differentiation program and the stimulation of estradiol production. Together these deficiencies lead to the complete arrest of folliculogenesis and the subsequent loss of fertility in animals lacking IGF1R in granulosa cells.

Abbreviations:

 
• Akt

  protein kinase B

 
• cAMP

  cyclic adenosine monophosphate

 
• eCG

  equine chorionic gonadotropin

 
• FSH

  follicle-stimulating hormone

 
• GC

  granulosa cell

 
• hCG

  human chorionic gonadotropin

 
• IGF

  insulinlike growth factor

 
• IGF1R

  insulinlike growth factor 1 receptor

 
• INSR

  insulin receptor

 
• LHR

  luteinizing hormone receptor

 
• PCR

  quantitative polymerase chain reaction

 
• SEM

  standard error of the mean.

Acknowledgments

This project was financed by National Institutes of Health (NIH) Grants R56HD086054 and R01HD057110 (to C.S.). S.C.B. was supported by NIH Training Grant T32HL07692.

Disclosure Summary: The authors have nothing to disclose.

Author notes

Address all correspondence and requests for reprints to: Carlos Stocco, PhD, 835 South Wolcott Avenue, Chicago, Illinois 60612. E-mail: [email protected].