Expression of Anti-Mullerian Hormone Protein during Early Follicular ...

0013-7227/07/$15.00/0 Printed in U.S.A.

Endocrinology 148(5):2273–2281 Copyright © 2007 by The Endocrine Society doi: 10.1210/en.2006-1501

Expression of Anti-Mullerian Hormone Protein during Early Follicular Development in the Primate Ovary in Vivo Is Influenced by Suppression of Gonadotropin Secretion and Inhibition of Vascular Endothelial Growth Factor Fiona H. Thomas, Evelyn E. Telfer, and Hamish M. Fraser Medical Research Council Human Reproductive Sciences Unit (F.H.T., H.M.F.), University of Edinburgh Centre for Reproductive Biology, The Queen’s Medical Research Institute, Edinburgh EH16 4TJ, United Kingdom; and Institute of Cell Biology (E.E.T.), The Darwin Building, University of Edinburgh, King’s Buildings, Edinburgh EH9 3JR, United Kingdom Anti-Mullerian hormone (AMH) plays a role during early follicular development and selection. The aim of this study was to determine the pattern of AMH protein expression in the marmoset ovary and to investigate the effects of inhibition of gonadotropins or vascular endothelial growth factor (VEGF) activity on AMH expression in vivo. GnRH antagonist or VEGF Trap, a soluble decoy receptor, was administered on d 0 or 5 of the follicular phase of the cycle, and ovaries were collected at the end of the follicular phase (d 10). AMH protein was expressed in the marmoset ovary in granulosa cells from the primary stage, with the most abundant staining at the preantral and early antral stages. Inhibition of gonadotropin secretion or VEGF activity between d 0 –10 of the cycle decreased AMH expression in early preantral follicles (P < 0.01),

A

NTI-MULLERIAN HORMONE (AMH), also known as Mullerian inhibiting substance (MIS), is a peptide growth factor and a member of the TGF-␤ superfamily of growth and differentiation factors (1). In the male, AMH is expressed in Sertoli cells from the fetal stage, whereas in the female, AMH is expressed postnatally in granulosa cells. AMH mRNA and protein are present in the granulosa cells of growing follicles in rats, mice, sheep, and humans (2–9). In rodent, human, and nonhuman primate ovaries, AMH is expressed mainly in granulosa cells of preantral and small antral follicles (6, 8, 10, 11) with expression declining in larger antral follicles. However, in contrast to rodents, AMH protein has been detected in human oocytes, stroma, and theca cells (9). The specific pattern of expression of AMH and the AMH receptor, AMHRII, in the ovary indicates a role for AMH during follicular development (6). In addition, AMH null mice provide insight into the role of AMH in follicular development (12), with AMH playing a role in the inhibition of recruitment of primordial follicles into the pool of growing First Published Online February 22, 2007 Abbreviations: AMH, Anti-Mullerian hormone; VEGF, vascular endothelial growth factor. Endocrinology is published monthly by The Endocrine Society (http:// www.endo-society.org), the foremost professional society serving the endocrine community.

and AMH expression was decreased in late preantral follicles in the presence of the VEGF Trap (P < 0.01), compared with controls. There was significantly less AMH expression in early antral follicles with both treatments (P < 0.01), and a decrease in the ratio of oocyte-associated/basement-membrane-associated granulosa cell expression of AMH (P < 0.05). When treatments were administered from d 5–10 of the cycle, both VEGF Trap and GnRH antagonist decreased AMH expression in preantral follicles (P < 0.01) but had no significant effect on early antral follicles. In conclusion, VEGF and gonadotropins are involved in the regulation of expression of AMH in the marmoset. This AMH expression may be a marker of abnormal folliculogenesis in the absence of gonadotropin stimulation or functional angiogenesis. (Endocrinology 148: 2273–2281, 2007)

follicles and decreasing the responsiveness of growing follicles to FSH (6). In human polycystic ovaries, AMH expression is increased due to the accumulation of small antral follicles (13), but these ovaries show a decrease in AMH expression in primordial and transitional follicles (9), which is associated with decreased granulosa cell number. These observations have led to the hypothesis that reduced AMH expression may be a marker of abnormal early follicular development and that AMH expression is associated with granulosa cell mitosis (2, 9). Surprisingly little information is available on the role of gonadotropins in the regulation of AMH expression. FSH has been shown to stimulate AMH transcription in a prepubertal Sertoli cell line (14), and treatment with recombinant FSH or estradiol results in down-regulation of AMH in early-stage follicles of prepubertal rats (5). GnRH antagonists are widely used during ovarian hyperstimulation programs to block endogenous gonadotropin secretion and the preovulatory LH surge (15). In the marmoset ovary, we have shown that reduction of gonadotropins by GnRH antagonist inhibits granulosa cell proliferation and selection of a dominant follicle (16), resulting in the accumulation of small antral follicles. To follow from the observations in human polycystic ovaries (9, 13), it would be timely to examine AMH expression in the ovaries of animals treated with GnRH antagonist. The ovary is one of the few organs in the body to undergo

2273

2274

Endocrinology, May 2007, 148(5):2273–2281

intense angiogenesis, and this process is involved in follicular selection, dominance and atresia (17). Vascular endothelial growth factor (VEGF) has been found to be the principal angiogenic factor during follicular development (17). The effect of inhibiting angiogenesis has been studied in the ovaries of the marmoset monkey using the VEGF Trap comprising Ig domain 2 of VEGF receptor type 1 and Ig domain 3 of VEGF receptor type 2, fused with human Fc (18). Through the inhibition of VEGF, these studies have established a requirement for VEGF during follicular vascularization, antrum formation, and ovulation (17, 18). In addition to these observations, proliferation of theca, endothelial, and granulosa cells is inhibited by the VEGF Trap (18). We have also previously shown that treatment of marmosets with the VEGF Trap or GnRH antagonist for the entire follicular phase (d 0 –10 of the cycle) results in the absence of dominant preovulatory follicles, whereas in the remaining antral follicles, granulosa, theca, and endothelial cell proliferation is reduced (16). By the midfollicular phase (d 5) in the marmoset, selection of the dominant follicles has already occurred (19). At this time, treatment with GnRH antagonist does not prevent preovulatory follicle development but increases atresia in the dominant follicles (16). Further characterization of these ovaries is required to determine the effects of these treatments on the expression of factors involved in early follicular development and selection. Recently, VEGF has been reported to stimulate preantral follicle development in the rat ovary during early follicular development (20), but little is known of the role of VEGF in regulating expression of putative intraovarian factors involved in follicular growth. There have been no previous reports of AMH protein expression in the nonhuman primate ovary, and little is known of the regulation of AMH expression during follicular development in vivo. In the present study, AMH expression will be analyzed in marmoset ovaries after inhibition of gonadotropin production or inhibition of VEGF in vivo. Materials and Methods Animals Adult female common marmoset monkeys (Callithrix jacchus), 2–3 yr old with a body weight of approximately 350 g and regular (28-d) ovulatory cycles, as determined by plasma progesterone concentrations of blood samples collected three times per week (21), were housed together with a younger sister or prepubertal female as described previously (22).

Treatments Experiments were carried out in accordance with the Animals (Scientific Procedures) Act, 1986, and were approved by the Local Ethical Review Process Committee. To synchronize follicular recruitment, selection, and ovulation during the treatment cycle and to render the length of the luteal phase similar to that of higher primates, marmosets were treated with 1 ␮g prostaglandin F2␣ analog (cloprostenol, Planate; Coopers Animal Health Ltd., Crewe, UK), im on d 13–15 of the 20-d luteal phase to induce luteolysis (19, 23). The day of prostaglandin injection was designated follicular d 0. This is associated with follicular recruitment, followed by selection on d 5 and ovulation around d 10 (23). To block the action of VEGF, we employed the VEGF Trap, a soluble decoy receptor created by fusing the extracellular domains of the human VEGF receptors 1 and 2 to the Fc portion of a human Ig. The incorporation of the Fc domain results in homodimerization of the recombinant

Thomas et al. • Regulation of AMH Expression in Vivo

protein, creating a high-affinity VEGF Trap (24). The VEGF trap binds all isoforms of VEGF-A, VEGF-B, and placental growth factor. The VEGF Trap was administered as a single sc injection at 25 mg/kg on d 0 (n ⫽ 4) or d 5 (n ⫽ 4) of the follicular phase as described previously (25), and ovaries were collected on d 10, corresponding to the periovulatory period in control animals. To suppress FSH and LH secretion from the pituitary (26 –28), the GnRH antagonist Antarelix (29) was used as described previously (16). Antarelix was injected at a dose of 12 mg/kg sc on follicular d 0 (n ⫽ 4) or d 5 (n ⫽ 4). As for the VEGF Trap group, ovaries were collected on d 10 of the follicular phase. Ovaries from control (vehicle) marmosets (experiment 1, n ⫽ 11; experiment 2, n ⫽ 4) were collected during the late follicular phase of the cycle. At the end of the treatment periods, animals were injected iv with 20 mg bromodeoxyuridine (Roche Molecular Biochemicals, Essex, UK) in saline 1 h before being sedated using 100 ␮l ketamine hydrochloride (Parke-Davis Veterinary, Pontypool, UK) and euthanized with an iv injection of 400 ␮l Euthetal (sodium pentobarbitone; Rhone Merieux, Harlow, UK). After cardiac exsanguination, ovaries were removed, weighed, and immediately fixed in 4% neutral buffered formalin. After 24 h, the ovaries were transferred to 70% ethanol, dehydrated, and embedded in paraffin according to standard procedures.

Immunohistochemistry Ovaries were embedded and serially sectioned, and tissue sections (5 ␮m) were placed onto BDH SuperFrost slides (BDH, Merck Co., Inc., Poole, UK). After dewaxing in xylene and rehydration in a series of ethanols, expression of AMH protein in the control and treated marmoset ovaries was detected by immunohistochemistry, using an open protocol on the Bond automated immunohistochemistry system (Vision BioSystems, Newcastle Upon Tyne, UK). After dewaxing in xylene and rehydration in ethanol, hydrogen peroxide (10% in methanol) was used for 30 min to block endogenous peroxide. Sections were then incubated for another 30 min with normal rabbit serum diluted 1:5 in Tris-buffered saline containing 5% BSA (Sigma Chemicals, Poole, UK) to block nonspecific binding sites, followed by blocking of endogenous avidin and biotin sites (avidin 15 min, biotin 15 min; Vector Labs, Peterborough, UK). The primary antibody to AMH (goat antihuman AMH; Santa Cruz Biotechnology, Santa Cruz, CA) was diluted 1:1000 in Bond antibody diluent, and incubation was carried out for 2 h at room temperature. For the negative controls, primary antibody was omitted. Incubation with the secondary antibody, biotinylated rabbit antigoat IgG [Vector Labs; 1:500 diluted in Bond antibody diluent (Vision BioSystems)] for 30 min was followed by avidin-biotin-horseradish peroxidase (Dako, Cambridgeshire, UK) for 30 min. The peroxidase reaction was developed with 3,3⬘-diaminobenzidine tetrachloride dihydrate (Dako). Finally, sections were counterstained with hematoxylin, dehydrated, and mounted. To assess cell death in the follicles of control and treated marmoset ovaries, immunohistochemistry for activated caspase-3 (Asp175; New England Biolabs, Hitchin, UK) was performed as described previously (25).

Quantification of immunohistochemistry Right and left ovaries from control animals (experiment 1, n ⫽ 11; experiment 2, n ⫽ 4), VEGF Trap-treated animals (d 0 –10, n ⫽ 4; d 5–10, n ⫽ 4), and GnRH antagonist-treated animals (d 0 –10, n ⫽ 4; d 5–10, n ⫽ 4) were examined. Stages of follicular development were defined as previously reported (18, 21), i.e. primary (oocyte surrounded by one granulosa cell layer), early secondary/early preantral (two to four granulosa cell layers, no antrum), late secondary/late preantral (more than four granulosa cell layers, no antrum), and tertiary/early antral (follicles containing an antrum, ⬍2 mm). Only those follicles with a visible oocyte containing a nucleus were considered to ensure proper follicular classification. AMH expression was detected using immunocytochemistry. Expression was analyzed in follicles from control ovaries and ovaries from marmosets treated with VEGF Trap or GnRH antagonist as described below.



2275

Experiment 1: relative AMH intensity during preantral follicular development

tion is based on previously described criteria for the measurement of follicular cell death (30).

Because AMH expression in other species has been reported to be most abundant in growing follicles, changes in AMH expression were initially investigated in early and late preantral follicles of marmoset control ovaries (n ⫽ 11 animals), with the intensity of AMH expression being classified using a visual scoring system (⫹, weak; ⫹⫹, moderate; ⫹⫹⫹, strong staining). Follicles within two ovaries per animal from the same immunohistochemistry run were analyzed; thus, changes in staining intensity between runs were eliminated. A total of 137 early preantral follicles and 45 late preantral follicles were analyzed. Additional experiments were subsequently performed to compare AMH expression in early preantral, late preantral, and early antral follicles using the criteria described in the following sections.

Statistical analyses

Experiment 2: AMH expression in preantral and early antral follicles of control ovaries and ovaries treated with VEGF Trap or GnRH antagonist For comparison of AMH expression in controls and treatment groups during follicular development (n ⫽ 4 animals per group), two methods of quantification were used. First, the presence or absence of AMH staining was noted in granulosa cells of early preantral, late preantral, and early antral follicles and expressed as a percentage of the total number of follicles counted in each group. Follicles were analyzed in three representative sections per ovary, and two ovaries per animal were used. The numbers of follicles analyzed were as follows: controls (early preantral n ⫽ 86; late preantral n ⫽ 29; early antral n ⫽ 18), VEGF Trap d 0 –10 (early preantral n ⫽ 166; late preantral n ⫽ 57; early antral n ⫽ 31), VEGF Trap d 5–10 (early preantral n ⫽ 106; late preantral n ⫽ 40; early antral n ⫽ 23), GnRH antagonist d 0 –10 (early preantral n ⫽ 113; late preantral n ⫽ 46; early antral n ⫽ 47), and GnRH antagonist d 5–10 (early preantral n ⫽ 101; late preantral n ⫽ 65; early antral n ⫽ 29). The second method used for analysis of AMH expression has been previously described for quantification of CD31 labeling in ovarian sections (18). This quantitative analysis was performed using an image analysis system linked to an Olympus camera, and the data processed using Image-Pro Plus for Windows (Microsoft). Briefly, the area of AMH-positive staining was measured at ⫻40 magnification in ovarian sections. The captured gray-scale image was thresholded and converted to a binary image. The whole area of the follicle (contained within the basement membrane) and the AMH-positive area within this compartment were measured. The AMH area was then calculated per unit area of the follicle and expressed as the mean value for the number of follicles assessed within each follicular stage. Follicles from two ovaries per animal were analyzed in the same immunohistochemistry run to eliminate any changes in staining intensity between runs. The numbers of follicles analyzed were as follows: controls (early preantral n ⫽ 57; late preantral n ⫽ 27; early antral n ⫽ 13), VEGF Trap d 0 –10 (early preantral n ⫽ 39; late preantral n ⫽ 38; early antral n ⫽ 12), VEGF Trap d 5–10 (early preantral n ⫽ 28; late preantral n ⫽ 24; early antral n ⫽ 12), GnRH antagonist d 0 –10 (early preantral n ⫽ 49; late preantral n ⫽ 21; early antral n ⫽ 11), and GnRH antagonist d 5–10 (early preantral n ⫽ 31; late preantral n ⫽ 26; early antral n ⫽ 13).

AMH expression in oocyte- and basement-membraneassociated granulosa cells In early antral follicles, additional analysis was carried out to determine the localization of AMH staining within each follicle. The area of granulosa cells immediately surrounding the oocyte was outlined, and the AMH-positive area within this compartment was measured and expressed as a percentage of the total area. Similarly, the area of AMHpositive cells in the compartment between the antrum and the basement membrane was calculated using the image analysis method described in the previous section.

Measurement of granulosa cell death Follicles from the groups described above were classed as atretic if they expressed protein for activated caspase-3 in more than 5% of their granulosa cells, as detected by immunohistochemistry. This classifica-

For quantification of the proportion of follicles expressing AMH, each follicle stage was compared between groups using ␹2 analysis, followed by Fisher’s exact test. For the quantitative image analysis, the area of AMH staining was expressed as a percentage of the total follicle area, and values were compared between treatments within each follicular stage using one-way ANOVA, followed by Tukey’s multiple comparison test (GraphPad Prism). P values ⬍ 0.05 were accepted as statistically significant.

Results AMH expression during follicular development

AMH protein was expressed in the marmoset ovary in granulosa cells from the primary stage, with abundant staining at the preantral and early antral stages, and decreased expression in late antral and atretic follicles. There was no AMH expression in the theca layer or in primordial follicles. Representative sections of ovaries from control animals (A) and animals treated with VEGF Trap (B) and GnRH antagonist (C) are shown in Fig. 1. In control ovaries, there was a higher percentage of follicles with no AMH/weak AMH staining at the early preantral stage (two to four granulosa layers) (P ⬍ 0.05) compared with the late preantral stage (more than four layers) (Fig. 2). There was an increase in the percentage of follicles with strong AMH staining at the late preantral stage (P ⬍ 0.05) (Fig. 2). In Fig. 3, representative sections showing lack of expression of AMH in primordial follicles (A), and AMH expression in primary (B), secondary (early preantral) (C), late preantral (D), and early antral follicles (E) are shown. Negative controls are shown in F–J. Effect of VEGF Trap and GnRH antagonist on the percentage of follicles showing positive AMH immunostaining

Figure 4 illustrates the effect of treatment with VEGF Trap or GnRH antagonist from d 0 –10 (A, C, and E) or d 5–10 (B, D, and F) on the percentage of follicles with positive AMH immunostaining in early preantral (A and B), late preantral (C and D), and early antral (E and F) follicles. Both treatments given from d 0 –10 of the cycle significantly decreased the percentage of early preantral follicles expressing AMH (P ⬍ 0.05) (A); however, the percentage of late preantral follicles expressing AMH was only decreased in the presence of the VEGF Trap (C). There was no effect on AMH expression in early antral follicles with either treatment given from d 0 –10 (E). When treatments were administered from d 5–10 of the cycle, the VEGF Trap decreased the percentage of early preantral follicles expressing AMH (P ⬍ 0.05) (B) but had no significant effect on late preantral or early antral follicles (D and F). There was no effect of GnRH antagonist when administered from d 5–10 of the cycle (B, D, and F). Effect of VEGF Trap and GnRH antagonist on the percentage of AMH immunostaining per follicle

Figure 5 illustrates the effect of treatment with VEGF Trap or GnRH antagonist from d 0 –10 (A, C, and E) or d 5–10 (B, D, and F) on the percentage of AMH expression (measured

2276



100

A

80 % Follicles

EPA

At

20

B

No AMH AMH+

y

40

LPA

0

EA

b

60

a

c

Early Preantral

z

AMH++ AMH+++

x Late Preantral

FIG. 2. AMH protein was detected in marmoset control ovaries by immunohistochemistry, and intensity of AMH expression was measured in early preantral and late preantral follicles using a visual scoring system (AMH⫹, AMH⫹⫹, AMH⫹⫹⫹) for weak, moderate, and strong staining, respectively. Different letters (a, x; b, y; and c, z) denote significant differences between the percentage of early and late preantral follicles expressing AMH (P ⬍ 0.05).

EPA

LPA

EPA

C

GnRH antagonist (d 0 –10) (C). When the treatments were given from d 5–10 of the cycle, AMH expression was decreased in late preantral follicles from both groups (P ⬍ 0.01) (D). Early antral follicles also had decreased AMH expression when VEGF Trap or GnRH antagonist was administered from d 0 –10 (P ⬍ 0.01) (E) but not when administered from d 5–10. AMH expression in oocyte- and basement-membraneassociated granulosa cells

EPA

FIG. 1. Representative ovarian sections showing AMH expression in control ovaries (A) and ovaries from animals treated with VEGF Trap (B) and GnRH antagonist (C). At, Atretic follicle; EA, early antral follicle; EPA, early preantral follicle; LPA, late preantral follicle. Bar, 300 ␮m.

by dividing the area of AMH immunostaining by the total follicle area) in early preantral (A and B), late preantral (C and D), and early antral (E and F) follicles. Unlike the quantification presented in Fig. 4, this analysis allowed changes in the levels of AMH per follicle to be ascertained. Both treatments given from d 0 –10 or d 5–10 of the cycle significantly decreased AMH expression in early preantral follicles (P ⬍ 0.01) (A and B). The expression of AMH in late preantral follicles was decreased in the presence of the VEGF Trap (d 0 –10) (P ⬍ 0.001), but there was no significant effect of the

Because early antral follicles displayed changes in AMH expression in the presence of the VEGF Trap and GnRH antagonist (treatment from d 0 –10), additional analysis was performed to determine any changes in the localization of AMH within each follicle. In controls, it was observed qualitatively that the granulosa cells adjacent to the oocyte had an increased area of AMH-positive cells compared with granulosa cells adjacent to the basement membrane. When quantified, the ratio of oocyte-associated/basement-membrane-associated AMH was decreased in the presence of the VEGF Trap and the GnRH antagonist (d 0 –10), compared with controls (Fig. 6A). Although overall AMH expression was not significantly different between the controls and the d 5–10 treatment groups, the ratio of oocyte-associated/basement membrane-associated AMH was decreased in the treated groups (P ⬍ 0.05) (Fig. 6B). Effect of VEGF Trap and GnRH antagonist on the expression of activated caspase-3

There was no granulosa cell death, as measured by activated caspase-3 expression, at the preantral stage of development in controls or ovaries treated with VEGF Trap or GnRH antagonist (d 0 –10 or d 5–10) (results not shown). In early antral follicles, there was no significant difference between groups in the percentage of follicles undergoing atresia, as measured by the expression of more than 5% activated caspase-3 in the granulosa cells (results not shown).


A

F

B

G

C

H

D

I

E

J

FIG. 3. Representative ovarian sections showing a lack of expression of AMH in primordial follicles (A) and positive AMH expression in primary (B, arrow), secondary (early preantral, C), late preantral (D), and early antral (E) follicles. F–J, Corresponding negative controls. Bar, 50 ␮m (A, B, F, and G) and 100 ␮m (C–E and H–J).

Discussion

In the marmoset ovary, AMH protein was present in granulosa cells from the primary stage, with the most abundant expression at the preantral and early antral stages, which is similar to the expression pattern reported in other species (6, 8, 10, 11). Here we demonstrate for the first time that AMH expression is decreased after inhibition of gonadotropin pro-


2277

duction or VEGF activity in vivo, suggesting a role for VEGF and gonadotropins in the regulation of AMH expression. We used two types of quantitative analysis of AMH immunostaining in follicles at the preantral and early antral stages of development. Interestingly, when the presence or absence of AMH staining was noted, the effect of inhibition of VEGF activity or gonadotropin production on AMH expression appeared to be stage specific, with AMH expression being decreased mainly in early preantral follicles. However, on further analysis, it became apparent that this method did not allow for changes in the extent of AMH expression within each follicle to be quantified. Therefore, we employed a second method of analysis that allowed the area of AMH staining per follicle to be quantified and expressed as a percentage of the total follicle area. In this way, changes in AMH expression in individual follicles due to the treatment groups could be analyzed in more detail. A differential pattern of expression in large preantral and early antral follicles has been previously described in the rat ovary (5), where follicles with lower AMH expression are thought to be more sensitive to FSH (12). Because the expression of AMH may affect FSH sensitivity of follicles, AMH may play a role in determining whether follicles undergo selection or atresia. In the present study, we have observed that AMH is expressed more extensively in the granulosa cells adjacent to the oocyte, compared with the granulosa cells close to the basement membrane. This pattern has previously been observed in the rat ovary (2, 5). A possible gradient of AMH expression within the follicle may reflect functional differences between granulosa cells surrounding the oocyte and the more peripheral granulosa cells, such as differences in proliferation capacity and steroidogenic activity (5). These functional differences arise under the influence of factors produced by the oocyte, such as growth/differentiation factor-9 (31). Baarends et al. (5) have also observed that expression of AMH often becomes restricted to the granulosa cells surrounding the oocyte in large preantral follicles and atretic follicles that are losing expression of AMH. We observed a similar pattern of expression in control ovaries. However, inhibition of VEGF or gonadotropin secretion in vivo results in a shift in the expression pattern of AMH within the early antral follicle, with a decrease in the ratio of oocyteassociated AMH/basement-membrane-associated AMH expression. This finding suggests that either the function of the granulosa cells or their response to oocyte-secreted factors may be affected by these treatments. The possibility that the treatments are altering granulosa cell differentiation is an interesting one. Because angiogenesis and gonadotropin stimulation are essential for normal follicular development, it is likely that disruption of these pathways will result in restriction in endocrine control as well as aberrant expression of local factors within the follicle. Moreover, because communication between the oocyte, follicular cells, and the extracellular matrix all contribute to follicular development, alterations in angiogenesis and gonadotropin stimulation may disrupt these processes. Additional studies are required to further elucidate the mechanisms of the findings reported here. The first study to evaluate the effects of AMH on intact ovarian follicles showed that AMH treatment enhanced preantral follicle growth in vitro by increasing both follicle size and cell number (32). This finding is consistent with the highest



A

B

100

100

80

80

*

60

*

40

% Follicles

% Follicles

2278

40 20

20

0

0 Control

VEGF Trap GnRH antag (0-10) (0-10)

C


Control


100

*

60

80

% Follicles

80

40 20

60 40 20

0

0 Control


E

F

100

100 % Follicles

80 % Follicles

Control

D

100

% Follicles

FIG. 4. Effect of treatment with VEGF Trap or GnRH antagonist (GnRH antag) between d 0 –10 (A, C, and E) or d 5–10 (B, D, and F) on the percentage of follicles expressing AMH in early preantral (A and B), late preantral (C and D), and early antral (E and F) follicles. *, Significantly different from control group (P ⬍ 0.05).

*

60

60 40 20

80 60 40 20 0

0 Control


expression of AMH in populations of cells that are most rapidly dividing (2, 33). In addition, AMH promoted follicle growth without enhancing differentiation (32), as demonstrated by the lack of regulation of inhibin-␣ subunit expression by AMH. This may have a long-lasting effect; when a follicle stimulated by AMH does eventually differentiate, it will be larger and contain more granulosa cells and thus have a greater capacity for producing estrogen and angiogenic factors, which in turn should provide a competitive advantage compared with other follicles in its cohort (32). Therefore, it may be expected that the effect of inhibition of angiogenesis or gonadotropin production in the present study would be to decrease AMH expression in growing follicles, because these treatments ultimately result in arrest of follicular development during the antral stage (16, 18). It should be noted, however, that another study has shown that AMH inhibits FSH-simulated development of murine preantral follicles in vitro (34). The possible explanation for this discrepancy is that that study used adult mouse ovaries, whereas McGee et al. (32) used prepubertal rats. A similar differential effect of TGF-␤ has been demonstrated, where TGF-␤ stimu-

Control


lates growth of preantral follicles derived from adults but not immature mice (35). The other possible explanation is that the mouse serum in the study by Durlinger et al. (34) may have contained factors influencing the action of AMH. There may be more than one explanation for the results of the current study. First, VEGF and gonadotropins may be involved in the regulation of AMH expression. An adequate vascular supply is necessary to provide endocrine and paracrine signals for the regulation of follicular development, and VEGF is one of the key factors regulating angiogenesis in the ovary. In the marmoset ovary, VEGF is produced in the theca and granulosa layers (16, 36), and inhibition of VEGF disrupts follicular growth and granulosa cell proliferation (18, 21). VEGF has also been reported to stimulate preantral follicle development in the rat ovary (20). This effect may be due to enhanced vascular development or vascular permeability in the proximity of developing follicles, resulting in increased delivery of endocrine or paracrine factors, or VEGF may have a direct mitogenic effect on granulosa cells, as has been shown in vitro (20). A role for VEGF in bovine granulosa cell survival has also been reported


A

B

10

10

8

8

*

6

*

4 2

% AMH expression

% AMH expression


4

VEGF Trap (0-10)

GnRH antag (0-10)

Control

D

12

10

10

8

% AMH expression

% AMH expression

C

8

*

6 4

VEGF Trap (5-10)

*

6

4

GnRH antag (5-10)

*

2

2 0

0

Control

VEGF Trap (0-10)

GnRH antag (0-10)

Control

VEGF Trap (5-10)

GnRH antag (5-10)

Control

VEGF Trap (5-10)

GnRH antag (5-10)

F

E

12

14

10

12

8

*

6 4

*

2

% AMH expression

% AMH expression

*

2 0

Control

FIG. 5. Effect of treatment with VEGF Trap or GnRH antagonist (GnRH antag) from d 0 –10 (A, C, and E) or d 5–10 (B, D, and F) on percentage of AMH expression per follicle area in early preantral (A and B), late preantral (C and D), and early antral (E and F) follicles. *, Significantly different from control group (P ⬍ 0.01).

*

6

0

2279

10 8 6 4 2 0

0 Control

VEGF Trap (0-10)

(37). Although primordial follicles do not have their own vascular supply (21), inhibition of VEGF signaling blocks primordial follicle activation in bovine ovarian cortical cultures (38). Because VEGF receptors are present in endothelial cells of the thecal layer from the secondary (preantral) stage of development in the marmoset (18), it is possible that VEGF indirectly stimulates AMH expression in the granulosa cells of follicles, which in turn inhibits activation of neighboring primordial follicles in a paracrine manner. In addition, because AMH has been shown to promote the additional growth of preantral follicles in rodents (32), by blocking VEGF activity in vivo and thus decreasing AMH expression, we can hypothesize a role for

GnRH antag (0-10)

VEGF in the promotion of follicular development via regulation of AMH expression in preantral follicles. However, whether the effect on granulosa cells is direct or indirect remains to be investigated. Few studies have investigated the role of gonadotropins in regulation of AMH expression in the ovary. In the rat ovary, treatment with GnRH antagonist in combination with recombinant FSH resulted in a decrease in AMH expression in some, but not all, preantral and small antral follicles (5). In contrast, in the male, administration of recombinant FSH to prepubertal mice results in an increase in testicular AMH output (39). Conversely, early studies have shown that ma-

2280


Oocyte/BM AMH

A


4 3

*

2

*

1 0 Control

Oocyte/BM AMH

B


4 3

*

2

*

1 0 Control


FIG. 6. Quantification of the ratio of oocyte-associated/basementmembrane-associated AMH (oocyte/BM AMH). *, Significantly different from control group (P ⬍ 0.05).

ternal treatment with antibody to GnRH causes an increase in AMH in postnatal testis, which could be reversed by FSH but not LH replacement (40), suggesting differing stage- and tissue-specific effects. To test further the effect of FSH, a subsequent study investigated the serum AMH levels in transgenic mice carrying a deletion in the gene encoding the FSH-␤ subunit (14). The results demonstrated a decrease in AMH levels in these mice, correlated with reduced Sertoli cell number and testicular volume. Similarly, our experiments using the GnRH antagonist support a possible stimulatory role for gonadotropins in the regulation of AMH protein expression in the primate ovary. An alternative explanation for the results presented here is that reduced AMH expression in the absence of VEGF activity or gonadotropin stimulation may be a marker for abnormal follicular development, evidenced by decreased proliferation or increased cell death. We did not observe any difference in the incidence of cell death between treatments; however, follicles may be directed to die without showing signs of atresia (41). Thus, the possibility that reduced AMH is an early signal of atresia in granulosa cells cannot be ruled out. Reduced AMH expression in preantral follicles treated with VEGF Trap or GnRH antagonist may also be a marker of abnormal folliculogenesis associated with a decrease in granulosa cell proliferation. This hypothesis is in agreement with a recent report by Stubbs et al. (9), who demonstrated that a decrease in AMH expression in early-stage follicles from polycystic ovaries was associated with a decrease in cell number. Previous work in our laboratory, using ovarian sections taken from the same animals treated with GnRH antagonist as the current study, has demonstrated an increase in granulosa cell proliferation in follicles

at the late preantral stage of development, compared with the early preantral stage (16). Here, we have shown that AMH is expressed in a higher proportion of follicles at the late preantral stage of development in control animals, suggesting a link between AMH expression and granulosa cell proliferation, as has previously been alluded to (2). However, we have failed to show that the decrease in AMH expression in early preantral follicles in the presence of the VEGF Trap or GnRH antagonist is associated with a decrease in granulosa cell proliferation, because our previous data using animals treated with the VEGF Trap or GnRH antagonist showed that there was no effect on proliferation until the early antral stage (16, 18). However, it is possible that early changes in AMH expression result in decreased granulosa cell proliferation later in development, i.e. once the antral stage has been reached. Alternatively, rather than decreased AMH expression being a marker of reduced proliferation, the effect of inhibition of angiogenesis and loss of gonadotropin stimulation may be a down-regulation of AMH expression in preantral and early antral follicles, either directly as a result of a shutdown of angiogenesis or gonadotropin stimulation or via a disruption in paracrine interactions between neighboring follicles. Our group has previously shown that treatment of marmosets with the VEGF Trap or GnRH antagonist for the entire follicular phase (d 0 –10 of the cycle) results in the absence of dominant preovulatory follicles, whereas in the remaining antral follicles, granulosa, theca, and endothelial cell proliferation is reduced (16). Alongside previous characterization of follicular development in animals treated with the VEGF Trap and GnRH antagonist, the results of the current study suggest that VEGF and gonadotropins may play a role in promoting follicular development during the preantral and early antral stages by influencing AMH expression. Interestingly, a decrease in AMH expression in the absence of angiogenesis or gonadotropin stimulation may actually increase FSH responsiveness, as has been reported in AMH knockout mice (34). This increase in responsiveness to gonadotropins after treatment with the VEGF Trap or GnRH antagonist may be responsible for the lack of effect of the treatments on granulosa cell proliferation and atresia during the early stages of development. When treatment of marmosets with the VEGF Trap or GnRH antagonist occurred before follicular selection (before d 5 of the follicular phase), the effects on AMH expression were more marked than when treatment was administered after follicle selection, particularly in early antral follicles. These results suggest that the effects of the VEGF Trap and GnRH antagonist on AMH expression in growing follicles may be dependent on the duration of treatment, the presence of an established vascular network, and the existing hormonal environment within the ovary. In conclusion, we have described the pattern of AMH protein expression during follicular development in the marmoset ovary and have demonstrated a role for VEGF and gonadotropins in the regulation of expression of AMH in growing follicles. Although reduced AMH expression during early follicular development was not associated with a decrease in granulosa cell proliferation or increased atresia, aberrant AMH expression may be an early marker of abnormal folliculogenesis in the absence of functional angiogenesis or gonadotropin stimulation.


Acknowledgments We thank Dr. J. S. Rudge, Dr. S. J. Weigand, and Regeneron Pharmaceuticals Inc. (Tarrytown, NY) for expert advice and gift of the VEGF Trap. We also thank Keith Morris and staff for animal care, Helen Wilson for help with histology, and Ian Swanston for the progesterone assays. Received November 9, 2006. Accepted February 13, 2007. Address all correspondence and requests for reprints to: Fiona H Thomas, Medical Research Council Human Reproductive Sciences Unit (F.H.T., H.M.F.), University of Edinburgh Centre for Reproductive Biology, The Queen’s Medical Research Institute, 47 Little France Crescent, Edinburgh, EH16 4TJ, United Kingdom. E-mail: [email protected]. Disclosure Statement: The authors have nothing to disclose.

References 1. Massague J 1990 The transforming growth factor-␤ family. Annu Rev Cell Biol 6:597– 641 2. Hirobe S, He WW, Lee MM, Donahoe PK 1992 Mullerian inhibiting substance messenger ribonucleic acid expression in granulosa and Sertoli cells coincides with their mitotic activity. Endocrinology 131:854 – 862 3. Munsterberg A, Lovell-Badge R 1991 Expression of the mouse anti-Mullerian hormone gene suggests a role in both male and female sexual differentiation. Development 113:613– 624 4. Ueno S, Takahashi M, Manganaro TF, Ragin RC, Donahoe PK 1989 Cellular localization of Mullerian inhibiting substance in the developing rat ovary. Endocrinology 124:1000 –1006 5. Baarends WM, Hoogerbrugge JW, Post M, Visser JA, De Rooij DG, Parvinen M, Themmen AP, Grootegoed JA 1995 Anti-Mullerian hormone and antiMullerian hormone type II receptor messenger ribonucleic acid expression during postnatal testis development and in the adult testis of the rat. Endocrinology 136:5614 –5622 6. Durlinger AL, Gruijters MJ, Kramer P, Karels B, Ingraham HA, Nachtigal MW, Uilenbroek JT, Grootegoed JA, Themmen AP 2002 Anti-Mullerian hormone inhibits initiation of primordial follicle growth in the mouse ovary. Endocrinology 143:1076 –1084 7. Bezard J, Vigier B, Tran D, Mauleon P, Josso N 1988 Anti-Mullerian hormone in sheep follicles. Reprod Nutr Dev 28:1105–1112 8. Weenen C, Laven JS, Von Bergh AR, Cranfield M, Groome NP, Visser JA, Kramer P, Fauser BC, Themmen AP 2004 Anti-Mullerian hormone expression pattern in the human ovary: potential implications for initial and cyclic follicle recruitment. Mol Hum Reprod 10:77– 83 9. Stubbs SA, Hardy K, Da Silva-Buttkus P, Stark J, Webber LJ, Flanagan AM, Themmen AP, Visser JA, Groome NP, Franks S 2005 Anti-Mullerian hormone protein expression is reduced during the initial stages of follicle development in human polycystic ovaries. J Clin Endocrinol Metab 90:5536 –5543 10. Rajpert-De Meyts E, Jorgensen N, Graem N, Muller J, Cate RL, Skakkebaek NE 1999 Expression of anti-Mullerian hormone during normal and pathological gonadal development: association with differentiation of Sertoli and granulosa cells. J Clin Endocrinol Metab 84:3836 –3844 11. Modi D, Bhartiya D, Puri C 2006 Developmental expression and cellular distribution of Mullerian inhibiting substance in the primate ovary. Reproduction 132:443– 453 12. Durlinger AL, Kramer P, Karels B, de Jong FH, Uilenbroek JT, Grootegoed JA, Themmen AP 1999 Control of primordial follicle recruitment by antiMullerian hormone in the mouse ovary. Endocrinology 140:5789 –5796 13. Pigny P, Jonard S, Robert Y, Dewailly D 2006 Serum anti-Mullerian hormone as a surrogate for antral follicle count for definition of the polycystic ovary syndrome. J Clin Endocrinol Metab 91:941–945 14. Lukas-Croisier C, Lasala C, Nicaud J, Bedecarras P, Kumar TR, Dutertre M, Matzuk MM, Picard JY, Josso N, Rey R 2003 Follicle-stimulating hormone increases testicular anti-Mullerian hormone (AMH) production through Sertoli cell proliferation and a nonclassical cyclic adenosine 5⬘-monophosphatemediated activation of the AMH Gene. Mol Endocrinol 17:550 –561 15. Huirne JA, van Loenen AC, Schats R, McDonnell J, Hompes PG, Schoemaker J, Homburg R, Lambalk CB 2004 Dose-finding study of daily gonadotropinreleasing hormone (GnRH) analogue for the prevention of premature luteinizing hormone surges in IVF/ICSI patients: antide and hormone levels. Hum Reprod 19:2206 –2215 16. Taylor PD, Hillier SG, Fraser HM 2004 Effects of GnRH antagonist treatment on follicular development and angiogenesis in the primate ovary. J Endocrinol 183:1–17 17. Fraser HM 2006 Regulation of the ovarian follicular vasculature. Reprod Biol Endocrinol 4:18


2281

18. Wulff C, Wilson H, Wiegand SJ, Rudge JS, Fraser HM 2002 Prevention of thecal angiogenesis, antral follicular growth, and ovulation in the primate by treatment with vascular endothelial growth factor Trap R1R2. Endocrinology 143:2797–2807 19. Gilchrist RB, Wicherek M, Heistermann M, Nayudu PL, Hodges JK 2001 Changes in follicle-stimulating hormone and follicle populations during the ovarian cycle of the common marmoset. Biol Reprod 64:127–135 20. Danforth DR, Arbogast LK, Ghosh S, Dickerman A, Rofagha R, Friedman CI 2003 Vascular endothelial growth factor stimulates preantral follicle growth in the rat ovary. Biol Reprod 68:736 –741 21. Wulff C, Wilson H, Rudge JS, Wiegand SJ, Lunn SF, Fraser HM 2001 Luteal angiogenesis: prevention and intervention by treatment with vascular endothelial growth factor trap(A40). J Clin Endocrinol Metab 86:3377–3386 22. Fraser HM, Dickson SE, Morris KD, Erickson GF, Lunn SF 1999 The effect of the angiogenesis inhibitor TNP-470 on luteal establishment and function in the primate. Hum Reprod 14:2054 –2060 23. Summers PM, Wennink CJ, Hodges JK 1985 Cloprostenol-induced luteolysis in the marmoset monkey (Callithrix jacchus). J Reprod Fertil 73:133–138 24. Holash J, Davis S, Papadopoulos N, Croll SD, Ho L, Russell M, Boland P, Leidich R, Hylton D, Burova E, Ioffe E, Huang T, Radziejewski C, Bailey K, Fandl JP, Daly T, Wiegand SJ, Yancopoulos GD, Rudge JS 2002 VEGF-Trap: a VEGF blocker with potent antitumor effects. Proc Natl Acad Sci USA 99: 11393–11398 25. Fraser HM, Wilson H, Wulff C, Rudge JS, Wiegand SJ 2006 Administration of vascular endothelial growth factor Trap during the ‘post-angiogenic’ period of the luteal phase causes rapid functional luteolysis and selective endothelial cell death in the marmoset. Reproduction 132:589 – 600 26. Fluker MR, Marshall LA, Monroe SE, Jaffe RB 1991 Variable ovarian response to gonadotropin-releasing hormone analogue-induced gonadotropin deprivation during different phases of the menstrual cycle. J Clin Endocrinol Metab 72:912–919 27. Hall JE, Bhatta N, Adams JM, Rivier JE, Vale WW, Crowley Jr WF 1991 Variable tolerance of the developing follicle and corpus luteum to gonadotropin-releasing hormone analogue-induced gonadotropin withdrawal in the human. J Clin Endocrinol Metab 72:993–1000 28. Fraser HM, Groome NP, McNeilly AS 1999 Follicle-stimulating hormoneinhibin B interactions during the follicular phase of the primate menstrual cycle revealed by gonadotropin-releasing hormone antagonist and antiestrogen treatment. J Clin Endocrinol Metab 84:1365–1369 29. Deghenghi R, Boutignon F, Wuthrich P, Lenaerts V 1993 Antarelix (EP 24332) a novel water soluble LHRH analogue. Biomed Pharmacother 47:107–110 30. Byskov AGS 1974 Cell kinetic studies of follicular atresia in the mouse ovary. J Reprod Fertil 37:277–285 31. Elvin JA, Clark AT, Wang P, Wolfman NM, Matzuk MM 1999 Paracrine actions of growth differentiation factor-9 in the mammalian ovary. Mol Endocrinol 13:1035–1048 32. McGee EA, Smith R, Spears N, Nachtigal MW, Ingraham H, Hsueh AJ 2001 Mullerian inhibitory substance induces growth of rat preantral ovarian follicles. Biol Reprod 64:293–298 33. Hirshfield AN, Schmidt WA 1987 Kinetic aspects of follicular development in the rat. Adv Exp Med Biol 219:211–236 34. Durlinger AL, Gruijters MJ, Kramer P, Karels B, Kumar TR, Matzuk MM, Rose UM, de Jong FH, Uilenbroek JT, Grootegoed JA, Themmen AP 2001 Anti-Mullerian hormone attenuates the effects of FSH on follicle development in the mouse ovary. Endocrinology 142:4891– 4899 35. Liu X, Andoh K, Abe Y, Kobayashi J, Yamada K, Mizunuma H, Ibuki Y 1999 A comparative study on transforming growth factor-␤ and activin A for preantral follicles from adult, immature, and diethylstilbestrol-primed immature mice. Endocrinology 140:2480 –2485 36. Fraser HM, Wulff C 2001 Angiogenesis in the primate ovary. Reprod Fertil Dev 13:557–566 37. Greenaway J, Connor K, Pedersen HG, Coomber BL, LaMarre J, Petrik J 2004 Vascular endothelial growth factor and its receptor, Flk-1/KDR, are cytoprotective in the extravascular compartment of the ovarian follicle. Endocrinology 145:2896–2905 38. Cushman RA, Allan MF, Clopton DT, Echternkamp SE, Cupp AS 2006 Inhibition of vascular endothelial growth factor signalling blocks primordial follicle activation in bovine ovarian cortical cultures. Biol Reprod Special Issue (Suppl 144; abstract 325) 39. Al-Attar L, Noel K, Dutertre M, Belville C, Forest MG, Burgoyne PS, Josso N, Rey R 1997 Hormonal and cellular regulation of Sertoli cell anti-Mullerian hormone production in the postnatal mouse. J Clin Invest 100:1335–1343 40. Bercu BB, Morikawa Y, Jackson IM, Donahoe PK 1979 Inhibition of Mullerian inhibiting substance secretion by FSH. Pediatr Res 13:246 –249 41. Durlinger AL, Kramer P, Karels B, Grootegoed JA, Uilenbroek JT, Themmen AP 2000 Apoptotic and proliferative changes during induced atresia of preovulatory follicles in the rat. Hum Reprod 15:2504 –2511

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.