REPRODUCTION RESEARCH

Transforming growth factor-b: its role in ovarian follicle development Davina Rosairo, Ileana Kuyznierewicz, Jock Findlay and Ann Drummond Prince Henry’s Institute of Medical Research, PO Box 5152, Clayton, Victoria 3168, Australia Correspondence should be addressed to A Drummond; Email: [email protected]

Abstract Ovarian follicular growth and differentiation in response to transforming growth factor-b (TGFB) was investigated using postnatal and immature ovarian models. TGFB ligand and receptor mRNAs were present in the rat ovary 4–12 days after birth and at day 25. In order to assess the impact of TGFB1 on follicle growth and transition from the primordial through to the primary and preantral stages of development, we established organ cultures with 4-day-old rat ovaries. After 10 days in culture with FSH, TGFB1, or a combination of the two, ovarian follicle numbers were counted and an assessment of atresia was undertaken using TUNEL. Preantral follicle numbers declined significantly when treated with the combination of FSH and TGFB1, consistent with our morphological appraisal suggesting an increase in atretic primary and preantral follicles. To investigate the mechanisms behind the actions of TGFB1, we isolated granulosa cells and treated them with FSH and TGFB1. Markers of proliferative, steroidogenic, and apoptotic capacity were measured by real-time PCR. Cyclin D2 mRNA expression by granulosa cells was significantly increased in response to the combination of FSH and TGFB. The expression of forkhead homolog in rhabdomyosarcoma (Foxo1) mRNA by granulosa cells was significantly reduced in the presence of both FSH and TGFB1, individually and in combination regimes. By contrast, the expression of steroidogenic enzymes/proteins was largely unaffected by TGFB1. These data suggest an inhibitory role for TGFB1 (in the presence of FSH) in follicle development and progression. Reproduction (2008) 136 799–809

Introduction Surprisingly little is known about the mechanisms involved in ovarian follicle development and what triggers some follicles to grow, differentiate and ultimately release a matured oocyte while others die. A group of structurally related proteins, known as the transforming growth factorb (TGFB) superfamily, have been implicated in the local regulation of ovarian function (for review see Knight & Glister 2006). Members of the family exert their effects via serine/threonine kinases that activate SMAD proteins. These receptor–SMAD complexes translocate to the nucleus where changes in nuclear transcription are elicited. Thus, for the effects of TGFB superfamily members to be mediated, all elements of the signalling pathway i.e. ligand, receptors, SMADs and any relevant co-factors must be present. We hypothesize that the combination of these components within a follicle changes during the transition of the follicle from primordial through primary, secondary and tertiary stages of development, and thus determines the biological effect of hormones and growth factors on the growth and differentiation of individual follicles. TGFB, the index member of the TGFB superfamily, comprises five subtypes, all products of separate genes, three of which, Tgfb1, Tgfb2 and Tgfb3, have been shown to be expressed in mammalian ovarian cells (Hernandez et al. 1990, Mulheron & Schomberg 1990, Derynck q 2008 Society for Reproduction and Fertility ISSN 1470–1626 (paper) 1741–7899 (online)

et al. 1998). These factors are synthesized in inactive precursor forms that undergo cleavage to produce monomers which can dimerize to 25 kDa forms through the conserved cysteine regions. The active regions of the TGFB monomers share 98–100% identity and functionally, TGFB1, TGFB2 and TGFB3 are indistinguishable in most bioassays (Knecht et al. 1987) giving rise to the suggestion that there is functional redundancy. Despite these findings, null mouse models for each of the TGFB ligands indicate non-overlapping phenotypes (Shull et al. 1992, Proetzel et al. 1995, Sanford et al. 1997) and therefore different physiological functions. In regard to fertility, these issues have been difficult to address given that about half of the TGFB1 null mice die during gestation with the remaining pups dying around the time of weaning and that TGFB2 and TGFB3 null mice exhibit perinatal lethality (for review see Dunker & Krieglstein 2000). However, if inflammatory pathologies can be contained, the lifespan of TGFB1 null mice can be extended so that they reach adulthood. Ingman et al. assessed TGFB1 null mice bred on the immunocompromised severe combined immunodeficiency spontaneous mutation (scid) background and established that their fertility was severely impaired. Irregular ovulation, a reduction in the number of oocytes ovulated and in their developmental competence was reported (Ingman et al. 2006). DOI: 10.1530/REP-08-0310 Online version via www.reproduction-online.org

800

D Rosairo and others

In the ovary, TGFB1 protein is expressed in granulosa cells of small bovine follicles (Nilsson et al. 2003) and in the large follicles of humans and mice (Chegini & Flanders 1992, Schmid et al. 1994, Ghiglieri et al. 1995, Christopher 2000). The theca of humans, mice, rats (Mulheron et al. 1991, Teerds & Dorrington 1992), pigs and hamsters (Roy et al. 1992, Teerds & Dorrington 1992, Roy & Hughes 1994, Levacher et al. 1996, May et al. 1996, Gueripel et al. 2004) express TGFB1 protein. In addition, mouse oocytes have been shown to express TGFB1 protein (Gueripel et al. 2004); whereas porcine granulosa cells express Tgfb1 mRNA but not protein (May et al. 1996). Granulosa and theca cells of bovine (Nilsson et al. 2003), human, rat and mouse (Mulheron et al. 1991) ovaries and mouse oocytes (Schmid et al. 1994, Gueripel et al. 2004) express Tgfb2 mRNA and protein. Whereas, Tgfb3 has been explored in a more limited way, the mRNA having been localized to theca and granulosa cells of follicles of all stages in mice (Schmid et al. 1994) and Tgfb3 mRNA and protein to bovine granulosa and theca cells (Nilsson et al. 2003). Tgfbr1 mRNA and protein are expressed by porcine granulosa cells (Goddard et al. 1995), whilst TGFBR1 protein is expressed by mouse luteal cells, granulosa cells, theca cells and oocytes (Juneja et al. 1996, Gueripel et al. 2004). Tgfbr2 mRNA is present in rat ovary (Tsuchida et al. 1993), whereas mRNA and protein have been localized to porcine granulosa cells (Goddard et al. 1995) and mouse theca, granulosa cells, oocytes and luteal cells (Schmid et al. 1994, Gueripel et al. 2004). In the human, TGFBR2 protein has been detected in granulosa, theca and interstitial cells (Roy & Kole 1998). Tgfbr3 (also known as b-glycan) mRNA and protein are present in porcine granulosa cells (Goddard et al. 1995) and we have previously reported that b-glycan mRNA is expressed by the ovaries of 4, 8 and 12 day old rats with localization of b-glycan protein to oocytes, granulosa cells and theca cells at all stages of folliculogenesis (Drummond et al. 2002). Despite the presence of TGFB ligands and receptors in the ovary of a range of species, the direct effects of the TGFB ligands on follicular development have not received much attention. In a study by Liu et al. (1999), the actions of TGFB1 on preantral follicular growth were investigated in vitro, on follicles isolated from immature and adult mice. Age-specific effects were recorded for follicle diameter with only preantral follicles from adult mice increasing in size. In this study, we investigated the role of TGFB in recruiting follicles into the growth pathway and for its capacity to mediate the transition of follicles from primary to preantral and preantral to antral stages of development. First, we determined the mRNA expression pattern of TGFB1, TGFB2, TGFB3, TGFBR1 and TGFBR2 in the rat ovary during postnatal development. We then took whole ovaries from 4-day-old rats, which contain only primordial and a few primary follicles and cultured them on floating Reproduction (2008) 136 799–809

filters for 10 days in the presence or absence of TGFB1 and FSH. The viability of the follicles was established by TUNEL and follicles at each developmental stage were counted. Since TGFB1 has been implicated in ovarian steroidogenesis and proliferation, we isolated granulosa cells from diethylstilboestrol (DES)-treated immature rats to investigate its effect on steroidogenic (side chain cleavage (CYP11A1), 3b-hydroxysteroid dehydrogenase (HSD3B) and STAR protein (STAR)) and proliferative (cyclin D2, forkhead homolog in rhabdomyosarcoma (FOXO1 or FOXO1A)) endpoints.

Results Expression of TGFB ligand and receptor mRNAs by rat ovaries TGFB1, TGFB2, TGFB3, TGFBR1 and TGFBR2 were present in rat ovaries as early as 4 days after birth (Figs 1 and 2 ). Expression of Tgfb1 mRNA increased twofold between days 8 and 12 before declining to day 8 levels at day 25. Tgfb2 mRNA declined between days 4 and 8, remaining low until day 12, when it increased to day 4 levels at day 25. Tgfb3 mRNA levels were similar to that of Tgfb2 (Fig. 1), the mRNA declining to a nadir at day 12 before assuming day 4 levels at day 25. mRNA for the type I and II TGFB receptors were differentially regulated with Tgfbr1 expression high at day 4, declining to a nadir at day 8 before increasing to recover day 4 levels at day 25. By contrast, Tgfbr2 appeared to be ubiquitously expressed during the post natal/immature period of development with no significant change in the levels of its mRNA (Fig. 2). Ovary organ cultures An initial time course study was undertaken to determine the most appropriate culture period (Fig. 3A–C and data not shown). Prior to culture, the ovaries of 4-day-old rats (Fig. 3A) contained 67% primordial follicles, 24% primary follicles and 9% preantral follicles (nZ8). Follicular development occurred with time in culture. After 5 days in culture (4C5 days), the ovaries contain primordial, primary and many preantral follicles, some with more than three layers of granulosa cells (data not shown). Thereafter, it was evident that centrally located follicles were being lost from the ovary most likely due to hypoxia and a lack of nutrients reaching the interior (Fig. 3B and C). Morphologically abnormal follicles were also present (Fig. 3C). Antral follicles were not observed in ovaries cultured up to 14 days in media alone (Fig. 3B and C). A culture time of 10 days was selected based on the morphological appearance of the ovary (Fig. 3D). Vasa staining clearly delineated the oocyte cytoplasm making it easier to count follicles (Fig. 3D–G). Primordial to preantral oocytes stained positively, independently of the treatment regimen. www.reproduction-online.org

TGFB action in the ovary

2.4

Tgfb1

a

a

1.6

8

b

a

801

Tgfbr1 ac

6

a c

0.8

4

0 1.5 Primer / Gapdh ratio

(n)

5

4

Tgfb2

a

a 1 b 0.5

b

0 5 3

4

6

8

(n)

b

0 3 2.5

Tgfb3

ac

a

6

(n)

Tgfbr2 a

2 a 1.5 a

0 5 b

4

1

0 4 5 D8 D12 Days post birth

8 D25

(n)

Figure 1 Levels of expression of TGFB ligands (mRNA) in postnatal (days 4, 8 and 12) and immature (day 25) rat ovaries, normalized for Gapdh mRNA expression. The number of pools (n) analysed separately at each age was 4–9. The data are meanGS.D., with different letters denoting statistical significance, P!0.05.

Follicle numbers The numbers and types of follicles present in ovaries after 10 days of culture in the presence of FSH, TGFB1 and combined treatment regimens are presented in Fig. 4A. Table 1 shows the proportion of follicles as a percentage of total follicle numbers. Ten days in culture without treatment had no significant effect on the proportion of primordial, primary or preantral follicles compared with freshly isolated 4-day-old ovaries (Fig. 4B and Table 1). Similar numbers of primordial and primary follicles were recorded across all treatment groups. Preantral follicle development, however, was found to be significantly reduced relative to untreated controls in ovaries cultured www.reproduction-online.org

3

0.5

abc

5 D4

4

1

a

2

2

9 Primer / Gapdh ratio

5

5

3

8

8 12 Days post birth

25

(n)

Figure 2 Levels of expression of TGFB receptors (mRNA) in postnatal (days 4, 8 and 12) and immature (day 25) rat ovaries, normalized for Gapdh mRNA expression. The number of pools (n) analysed separately at each age is indicated below each histogram. The data are meanGS.D., with different letters denoting statistical significance, P!0.05.

with FSH and TGFB1 (PZ0.017; Fig. 4A and B), but was not different to the number of preantral follicles in 4-dayold fresh (uncultured) ovaries (Fig. 4B). Apoptosis To assess the degree to which follicular apoptosis might be affected by our treatment regimen, TUNEL staining was implemented (Fig. 5). Very little apoptosis was observed in day 4 ovaries prior to culture (Fig. 5A). While it was not our intention to undertake a quantitative assessment, we have attempted to gauge the proportion of follicles that were atretic/healthy as determined by TUNEL staining (Fig. 6). After 10 days in culture, there was an increase in the number of apoptotic follicles present in the ovaries of all treatment groups relative to the uncultured 4-day-old ovary (Fig. 5A–E). The pattern of atresia across follicle types and treatments was similar (more healthy than atretic follicles) except for the combined FSHCTGFB1 group where there appeared to be more atretic primary and preantral follicles (Fig. 6) than healthy follicles. There was also the suggestion that FSH and TGFB1 individually reduced Reproduction (2008) 136 799–809

802

D Rosairo and others

A

B

C

AF Vt p

50.0 µm

200 µm

200 µm

D

E

Pr 200 µm

nu Oo cy 50.0 µm

F

50.0 µm

Figure 3 Histological appearance of ovaries before and after whole ovary organ culture for (A) 4-dayold ovary prior to culture, (B) 14 days, (C) 14 days high power showing abnormal follicles. (D–G) Vasa staining of oocytes in ovaries before and 10 days after whole ovary organ culture with treatment. (D) Control (inset showing immunohistochemical blank control), (E) FSH, (F) TGFB1, (G) combined FSH/TGFB1. A–B have scale bars of 200 mm, while C–G are set at 50 mm. AF, abnormal follicle; cy, cytoplasm; nu, nucleus; Oo, oocyte; Pr, preantral; P, primary; p, primordial; Vt, vacant tissue.

G

P

50.0 µm

The expression of cyclin D2 mRNA by cultured granulosa cells was not affected by FSH or TGFB1 individually, but in combination a significant increase was noted (Fig. 7). Foxo1 mRNA was significantly reduced in the presence of both FSH and TGFB1, individually and in combination regimes (Fig. 7). Cyp11a1 mRNA expression was essentially unchanged by either FSH or TGFB1 alone or in combination, whereas Star mRNA expression was enhanced by FSH, but not by TGFB1 (Fig. 8). Hsd3b mRNA expression was significantly inhibited by FSH and slightly elevated by TGFB1, whereas the combined FSH/ TGFB1 appeared to mimic the FSH response (Fig. 8).

Discussion Members of the TGFB superfamily, notably GDF9 and -9B and AMH, have been shown to influence the growth of early stage follicles (for review see Knight & Glister 2006). Few studies, however, have addressed the role that TGFB itself might play in the growth and Reproduction (2008) 136 799–809

A Follicle number

Effects of TGFB1 on the mRNA expression of steroidogenic proteins and mediators of proliferation

development of ovarian follicle populations. In these studies, we have chosen to use our established model of postnatal folliculogenesis (Drummond et al. 2002) to investigate follicle development and transition, as it occurs for the first time, in response to TGFB. PCR studies

800 600 400 200 *

0 Control

FSH (100 ng/ml)

TGFB1(10 ng/ml)

Treatment

FSH+TGFB1 Primordial Primary Preantral

B 140 Follicle number

atresia in the primordial follicle population, whereas FSH alone appeared to reduce the number of healthy preantral follicles (Fig. 6).

50.0 µm

100 60 * 20 0 Day 4 (n =8)

Control (n =6) FSH+TGFB1(n =6)

Fresh ovary

Day 4 ovaries cultured for 10 days

Figure 4 (A) Numbers of primordial, primary and preantral follicles in ovaries cultured for 10 days with either: FSH, TGFB1 or combined FSH/TGFB1. (B) Numbers of preantral follicles in uncultured 4-day-old ovaries and control and FSH/TGFB1 treated ovaries cultured for 10 days. The data are meanGS.D., nZat least six ovaries. *P!0.05 compared with preantral follicle control. www.reproduction-online.org

TGFB action in the ovary Table 1 Proportion of follicle types represented as a percentage of total follicle numbers, before (day 4) and after 10 days in culture with treatment. Whole ovary culture

Primordial Primary Preantral

Day 4 (%)

Control (%)

FSH (%)

TGFB1 (%)

FSH/TGFB1 (%)

67 24 9

60 31 10

67 29 5

75 22 6

73 23 4

on whole postnatal/immature ovaries established that the TGFB ligands (TGFB1–3) and receptors (TGFBR1 and 2) were present during the developmental period. While it was not our intention to localize ligands and receptors to follicles or cell types, or to quantitate their expression in this context, we can draw some

A

803

conclusions based on our knowledge of the follicle populations present in the ovary during this developmental period. Tgfb2 and Tgfbr1 mRNA expression is high in the ovary 4 days after birth when there are mainly primordial and primary follicles and precursor cells present. Between days 4 and 8, these mRNAs decline as primary and preantral follicles increase in prevalence and granulosa cells multiply. Tgfb2 mRNA remains low at day 12, suggesting that it is the cells of early growing follicles that express this mRNA. Between days 8 and 12, Tgfbr1 mRNA increases coinciding with the differentiation and recruitment of theca cells, the proliferation of granulosa cells and the formation of antral follicles. By contrast, Tgfbr2 mRNA appears to be ubiquitously expressed in the ovary during the postnatal/immature developmental period. We have previously reported that b-glycan (Tgfbr3) and the appropriate SMADs are

B Pr

P

100 µm

C

50.0 µm

D

50.0 µm

50.0 µm

E

F

St

p

50.0 µm

G

50.0 µm

H p

Pr

50.0 µm

www.reproduction-online.org

50.0 µm

Figure 5 TUNEL staining in ovaries before and after 10 days of whole ovary culture. (A) Four-day-old ovary, (B) Control, (C) FSH, (D) TGFB1 and (E) combined FSH/TGFB1 after 10 days of culture, immature testis (F) (positive control), (G) and (H) TUNEL negative control sections. The length of the scale bars in B–G is 50 mm, while in A and H it is 100 mm. P, primary; p, primordial; Pr, preantral; St, seminiferous tubule, arrow heads show granulosa cells and arrows positively stained oocytes. Reproduction (2008) 136 799–809

804

D Rosairo and others

74%

70%

75

Healthy Atretic 30%

50

26%

25%

25 0 FSH 100

93% 73%

75

Percentage

50

50%

b

0.28

b b

0.21

b b

0.14 0.07 0.00

50%

7%

0

100 75 50 25 0

a

27%

25

125 100 75 50 25 0

0.35

Foxo1 / Gapdh ratio

Control 75%

TGFB1 95%

88% 62% 38% 12% 5%

FSH / TGFB1 75%

62% 72% 25%

Primordial

38%

28%

Primary Follicle class

Preantral

Figure 6 Percentage of atretic (dark histograms) versus healthy (open histograms) follicles in ovaries cultured for 10 days with media, FSH, TGFB1 and FSH/TGFB1. The data are presented as meanGS.D.

present in the ovary at this time (Drummond et al. 2000). These results indicate that TGFB signals can be transduced and that it is biologically possible for TGFB to influence folliculogenesis. Since changes in follicle growth in response to TGFB1 have been demonstrated previously (Liu et al. 1999), the remaining studies were undertaken with this ligand and not TGFB2 or TGFB3. Follicle development in response to TGFB1 (in the presence or absence of FSH) was evaluated in an in vitro organ culture system. The results of our initial time course study indicated that a culture period of 10 days was optimal for 4-day-old rat ovaries (age at the start of culture). The ovaries remained morphologically healthy and follicle development occurred spontaneously with primordial, primary and preantral follicles evident in the ovary at the end of the culture period. No antral follicles, however, were observed. Continued culture after day 10 resulted in a loss of follicles as indicated by areas of vacant ovarian interstitial tissue and morphologically atretic follicles predominantly in the central regions of the ovary, indicative of a lack of oxygen and nutrients diffusing through to the interior. With this model, we wanted to determine whether the number of follicles in the ovary changed; if progression between the developmental stages was altered in any way and if the viability of the follicles was affected. We found that combined FSH and TGFB1 treatment, reduced the number of preantral follicles present in whole ovaries cultured for Reproduction (2008) 136 799–809

Cyclin D2 / Gapdh ratio

100

b

1.00

b

0.75 0.50

a

a

a

a

0.25 0.00

1 ng 10 ng Control FSH 100 ng/ml TGFB1 ng/ml

1 ng 10 ng FSH (100 ng/ml) + TGFB1

Figure 7 The expression of cyclin D2 and Foxo1 mRNA by granulosa cells isolated from DES-treated immature rats, in response to FSH, TGFB1, or FSH/TGFB1 (corrected for Gapdh). The data are presented as meanGS.D., nZ3–4 and are representative of three separate experiments. Different letters denote statistical significance, P!0.05.

10 days. Our TUNEL assessments support this data and suggest an increased apoptosis in primary and preantral follicle populations. Thus, it would appear that fewer follicles successfully make the transition to the preantral follicle stage in the presence of FSH and TGFB1. TGFB has been shown to increase the expression of FSH receptor mRNA and to prolong its stability, an effect that is enhanced in the presence of FSH (Inoue et al. 2003), but it is unlikely that this is the mechanism of action given that primordial and primary follicles are unresponsive to FSH stimulation. Since we have access to a well-characterized granulosa cell culture system, we decided to investigate granulosa cell functions that might be mediated by FSH and TGFB1. There are, however, limitations to this model in relation to the whole ovary system. The granulosa cells were isolated from immature rats treated with oestrogen (DES) for 4 days. These granulosa cells have undergone significant proliferation before being cultured and are responsive to FSH. Basically, they are more representative of preantral rather than primary or primordial follicle populations. Using a similar model, Liu et al. (1999) isolated preantral follicles from the ovaries of DES-treated 28-day-old mice and reported that FSH and TGFB1 individually increased follicular diameter and therefore growth, but in combination follicular diameter was reduced. These observations are consistent with our findings in whole ovary cultures that we report here. www.reproduction-online.org

TGFB action in the ovary

Hsd3b / Gapdh ratio

0.20 c

c

0.15 a 0.10

b

b

b

0.05 0.00

Star / Gapdh ratio

5

b b

4 b 3 2 1

a

a

a

0 ab

Cyp11a1 / Gapdh ratio

0.12 ab 0.09 a

ac

a

0.06

ac 0.03 0.00

Control

FSH 1 ng 10 ng 100 ng/ml TGFB1 ng/ml

1 ng 10 ng FSH (100 ng/ml) + TGFB1

Figure 8 The expression of Cyp11a1, Star and Hsd3b mRNA by granulosa cells isolated from DES treated immature rats, in response to FSH, TGFB1 or FSH/TGFB1 (corrected for Gapdh). The data are presented as meanGS.D., nZ3–4 and are representative of three separate experiments. Different letters denote statistical significance, P!0.05.

FOXO1 regulates Fas ligand (pro apoptotic factor; Brunet et al. 1999) and p27KIP (cell cycle inhibitor) expression. We were able to confirm the findings of Richards et al. (2002) that FOXO1 expression in granulosa cells is downregulated by FSH, suggesting that apoptosis might be reduced in these cells. Consistent with this observation is the functional luteinization that occurs in vitro when granulosa cells are treated with FSH. Differentiation of granulosa cells is associated with reduced apoptosis (Song et al. 1999). In a seemingly paradoxical situation, FOXO1 levels are elevated in the granulosa cells of growing follicles that express high levels of cyclin D2, an important cell cycle mediator (Ingman et al. 2006). While these observations appear to be at odds, it is likely that FOXO1 is linked to both the proliferative and apoptotic pathways of granulosa cells and that it is the stage of follicle development and their response to hormones that determines the role of FOXO1 at any given time. Primordial and primary follicles do not respond to FSH www.reproduction-online.org

805

stimulation and apoptosis in these populations is normally low. These observations suggest that alternative mechanisms regulate growth and differentiation in these follicle populations. Early preantral follicles, however, are just beginning to respond to FSH and contain granulosa cells at the start of a period of exponential growth. These follicles initially may express high levels of FOXO1 but with continued growth these levels are likely to decline consistent with the increase in apoptosis observed in growing follicle populations. Cyclin D2 mRNA expression by granulosa cells was enhanced by TGFB1, but only when FSH was present. There is some debate as to whether FSH itself stimulates cyclin D2 expression by granulosa cells. In our own studies, we have found an effect of FSH to be sporadic (this study and unpublished observations), while others (Ogawa et al. 2003) reported no effect and some a stimulation of cyclin D2 expression (Ingman et al. 2006). FSH has been shown to activate the PI3-kinase pathway which leads to inactivation of FOXO1 by AKT-dependent phosphorylation (Richards et al. 2002) effectively releasing the FOXO1-induced repression of cyclin D2 (Park et al. 2005). This, however, does not seem to be enough to stimulate granulosa cell proliferation in response to FSH stimulation in vitro. The combined effects of FSH and activin have been shown to stimulate granulosa cell proliferation in vitro (Ogawa et al. 2003) leading Park et al. (2005) to hypothesize that cross talk in granulosa cells between FOXO1 and the phosphorylated 2/3SMADs results in FOXO1 being antagonized. Since TGFB also acts via SMADs 2/3, a similar mechanism can be envisaged, although in an inhibitory context. Further studies are required to dissect the mechanisms by which individual follicle populations grow and differentiate, but here we provide evidence to support an inhibitory role for TGFB1 in preantral follicle growth that most likely involves an increase in apoptosis at the primary and preantral stages of follicle development.

Materials and Methods Animals Sprague–Dawley rats were obtained from Central Animal Services, Monash University (Melbourne, Australia). Ovaries were collected from untreated rats at 4, 8, 12 and 25 days of age. Ovaries were used either for organ culture, RNA extraction or for the preparation of formalin-fixed, paraffin-embedded tissue blocks. In addition, some animals at 21 days of age received a DES implant for 96 h, prior to ovary collection and the isolation of granulosa cells (Drummond et al. 1999). Animals were maintained under standard conditions of lighting and temperature and received laboratory feed pellets and water ad libitum. The project was approved by the institutional Animal Experimentation and Ethics Committee as conforming to the guidelines of the National Health and Medical Research Council of Australia. Reproduction (2008) 136 799–809

806

D Rosairo and others

RNA extraction

enzyme, dNTPs, reaction buffer and SYBR GREEN I dye were supplied in the FastStart DNA Master SYBR Green I kit (Roche); after mixing the reagents 2 ml/capillary was added. Primer concentrations of 10 pmol and 3 mM Mg were added to each capillary. Primer sequences are given in Table 2. The capillary volume was made up to 20 ml with sterile water. Forty cycles of PCR were programmed to ensure that the threshold crossing point (cycle number) was attained. Fluorescence emission was monitored continuously during cycling. At the completion of cycling, melting curve analysis was carried out to establish the specificity of the amplicons produced. In addition, each amplicon was sequenced to verify the identity of the amplified product (data not shown). The level of expression of each mRNA and their estimated crossing points in each sample were determined relative to the standard preparation using the LightCycler computer software. A ratio of specific mRNA/ housekeeper gene amplification was then calculated. In most instances, individual pools for each age group or treatment with each primer set were performed in a single PCR experiment. The intra-assay variation was never more than 5% (nZ7) regardless of the primer set. The nature of Lightcycler PCR diminishes issues such as assay sensitivity and, at the concentrations of standard and sample utilized in these studies, the sensitivity threshold (picograms) was never approached.

Ovaries were dissected free of fat and adhering tissue and homogenized in 1 ml Ultraspec RNA reagent (Biotecx: Fisher Biotec, Melbourne, Australia). After 5 min on ice, 0.2 ml chloroform per ml of Ultraspec RNA reagent was added to the samples, which were then shaken vigorously and stored at 4 8C for 5 min prior to centrifuging for 15 min at 12 000 g. RNA was precipitated from the aqueous phase with 1 vol of isopropanol, after which the pellet was washed twice with ethanol, air-dried and resuspended in sterile water. To ensure that the RNA was completely dissolved, the samples were incubated for 10 min at 60 8C. The samples were then treated with DNA free (Ambion: Austin, TX, USA) to remove DNAses. At least three independent pools of RNA were prepared for each age/treatment group. The number of ovaries/pool ranged from 24 to 40, for postnatal animals and 2/pool for immature animals.

Reverse transcription RNA (1 mg) was reverse transcribed with 50 units MMLV (Expand) reverse transcriptase (Roche) and final concentrations of 1! cDNA synthesis buffer (supplied with enzyme), 1 mM NTPs (Roche), 20 units Rnasin (Promega) 10 mM dithiothreitol and 25 pmol oligo dT15 (Roche), as previously described (Drummond et al. 1999).

Granulosa cell cultures Real time PCR

Granulosa cells were released from 25-day-old, DES-treated rat ovaries by repeated puncture with fine gauge needles, as previously described (Xiao et al. 1992). After washing and counting, the granulosa cells (5!105/well) were plated and incubated for 24 h at 37 8C in McCoys 5C (containing glutamine 2 mM, transferrin 100 mg/ml and penicillin 100 U/ml,

mRNA expression was analysed using the Roche LightCycler (Roche) as previously described (Drummond et al. 2000). Briefly, an ovarian cDNA pool diluted 1:2–1:2000 was used as the standard for the analyses. Sample cDNAs diluted 1:2–1:10 in sterile water were added to individual capillaries. Taq

Table 2 Oligonucleotide primer sequences used to amplify transforming growth factor-b (Tgfb)1–3, Tgfbr1 and Tgfbr2, Cyp11a1, Star, 3bhydroxysteroid dehydrogenase (Hsd3b), cyclin D2, forkhead homolog in rhabdomyosarcoma (Foxo1) and Gapdh cDNAs. The GenBank accession number, amplified product sizes and the 5 0 nucleotide (in parenthesis) are given. Where a rat sequence was unavailable primers were designed using the mouse sequence. Primer Tgfb1 Tgfb2 Tgfb3 Tgfbr1 Tgfbr2 Star Cyp11a1 Hsd3b Cyclin D2 Foxo1 Gapdh

Sequence 0

Product size (bp) 0

5 -GCT AAT GGT GGA CCG CAA CAA C-3 (745) 5 0 -CAG CAG CCG GTT ACC AAG-3 0 (970) 5 0 -TTG GAT GCC GCC TAT TGC T-3 0 (1301) 5 0 -CTG TTC GAT CTT GGG CGT ATT G-3 0 (1597) 5 0 -ATG CAC TTG CAA AGG GCT CT-3 0 (611) 5 0 -CCC TGG ATC ATG TCG AAT TTA TGG-3 0 (942) 5 0 -GCT GAC ATC TAT GCA ATG GG-3 0 (1202) 5 0 -ATA TTT GGC CTT AAC TTC TGT TC-3 0 (1368) 5 0 -CCA GGG CAT CCA GAT CGT GTG-3 0 (1772) 5 0 -TAG TGT TCA GGG AGC CGT CTT-3 0 (1947) 5 0 -GTG TCA TCA GAG CTG AAC ACG G-3 0 (719) 5 0 -GGC TGG CGA ACT CTA TCT GG-3 0 (845) 5 0 -TAC TTG GGC TTT GGC TGG GGT GTT-3 0 (1387) 5 0 -GGC AGG TAA TCA CAG AGT GCT GTT-3 0 (1669) 5 0 -GGC AAA TTC TCC ATA GCC AA-3 0 (872) 5 0 -GAA GGC AAG CCA GTA GAG C-3 0 (1088) 5 0 -CAT TGA GCA CAT CCT ACG CAA-3 0 (654) 5 0 -CAT TCA CTT CCT CGT CCT GCT-3 0 (821) 5 0 -TCA AGG ATA AGG GCG ACA GC-3 0 (68) 5 0 -CCC GGA CTG GAG AGA TGC TT-3 0 (299) 5 0 -GAC CCC TTC ATT GAC CTC AAC-3 0 (163) 5 0 -GAT GAC CTT GCC CAC AGC CTT-3 0 (703)

Reproduction (2008) 136 799–809

Accession no.

229

NM021578

297

NM031131

332

NM009368

167

NM009370

176

NM031132

168

BC088859

329

J05156

243

BC089937

188

D16308

251

AF247812

561

M32599

www.reproduction-online.org

TGFB action in the ovary

streptomycin 100 mg/ml, and fungizone 250 ng/ml). The next day the media was changed and treatments were added for 2 h; FSH (100 ng/ml; rFSH-I8, obtained from the National Hormone and Pituitary Distribution Program and the NIADDK, NIH, Baltimore, MD, USA) and or TGFB1 (1 and 10 ng/ml). The doses were based on previous studies undertaken in our laboratory. At the end of the incubation period, RNA was extracted as described.

Whole ovary cultures Whole ovaries were cultured using a protocol similar to that described by Nilsson et al. (2001). Ovaries from 4-day-old rats were dissected free of the ovarian bursa and other extraneous tissue. Whole ovaries were cultured for 10 days on floating filters (0.4 mm Millicell-CM; Millipore Corp., Bedford, MA, USA) in 0.5 ml Dulbecco-modified Eagle medium (DMEM)– Ham F-12 medium (1:1, v/v) containing 0.1% BSA (Sigma), 0.1% Albumax II (Gibco BRL), 5! ITS-X (supplement containing insulin, sodium, transferrin, sodium selenite, ethanolamine; Life Technologies, Inc.), and 0.05 mg/ml L-ascorbic acid (Sigma) in a 24-well culture plate, each with one control and one treatment group (Nunc plate, Applied Scientific, South San Francisco, CA, USA). Medium was supplemented with penicillin and streptomycin to prevent bacterial contamination. Ovaries were randomly assigned to control or treatment groups with three ovaries (from three separate rats) per floating filter in a 30 ml drop of treated DMEM/Ham’s F-12 medium. The ovaries were cultured in media alone (control), or media containing TGFB1 (10 ng/ml: R&D systems, Minneapolis, MN, USA), FSH (100 ng/ml), or a combination of TGFB1 and FSH. Duplicate filters were used for each treatment group (nZ6 ovaries). Ovaries were cultured at 37 8C in a humidified atmosphere containing 5% CO2 for 10 days. Treated DMEM/Ham’s F-12 media was replaced every 2 days. At the end of the culture period, ovaries were fixed in 10% formalin for 5 h, embedded in paraffin and sectioned at 5 mm. Some sections were stained with haematoxylin for morphological assessment, while others were used for immunohistochemistry/follicle counting and TUNEL staining.

Immunohistochemistry Vasa-mouse homologue (Abcam, Cambridge, UK), an oocytespecific marker, was localized to rat oocytes in the primordial stage of development, to aid follicle counting. Sections (5 mm) of organ-cultured, formalin-fixed paraffin-embedded rat ovary were stained using standard immunohistochemical protocols. Briefly, sections were dewaxed in histosol (Australian Biostain, Traralgon, VIC, Australia), dehydrated in ethanol and washed in water. Sections were placed in citrate buffer (0.1 M, pH 6) and microwaved (900 W) for 10 min to retrieve antigens. The sections were cooled and then equilibrated in 0.1 M PBS pH 7.4. Endogenous peroxidase activity was blocked by incubating the sections for 30 min in 0.3% hydrogen peroxide followed by three washes in distilled water. The sections were blocked for 30 min in 1% blocking reagent (Roche), after which the primary antibody diluted 1:800 was added and the sections incubated for 18 h at 4 8C. Following extensive washing in PBS, the biotinylated www.reproduction-online.org

807

second antibody diluted 1:200 (Dako, Sydney, Australia) was added to the sections for a 60 min incubation at room temperature. After washing in PBS, the sections were incubated with a peroxidase conjugated avidin–biotin complex (Vector Elite, Vector Labs, Burlingame, CA, USA) for 60 min at room temperature after which the reaction product was developed using 3,3 0 diaminobenzidine tetrahydrochloride (DAB; Dako) and hydrogen peroxide in PBS. The sections were counterstained with haematoxylin, dehydrated in ethanol, cleared in histosol and coverslips mounted using DPX (BDH, VWR International Ltd, Poole, UK). Control sections received buffer in place of primary antibody. Follicles were classified as previously described (Drummond et al. 2002).

Follicle counting Cultured ovaries were fixed, paraffin embedded and sectioned at 5 mm. Three serial sections were mounted/slide and every alternate slide (first and third sections) was used for Vasa staining and subsequent counting. Primordial, primary and preantral follicles were counted. Images were captured using a !20 objective and only Vasa-positive oocytes (follicles) in which a nucleolus could be visualized, were counted using analySIS Professional Imaging software, version 5.0 (Imaging Research Inc., Ontario, Canada). Percentage distribution was determined for each follicle class in a given treatment group.

TUNEL staining Apoptotic cells in ovarian sections were detected using terminal deoxynucleotidyl transferase (TdT) mediated dUTPbiotin nick-end labelling (TUNEL). The ApopTag Peroxidase in situ apoptosis detection kit (Chemicon International, Melbourne, Australia) was utilized. Ovarian sections mounted on superfrost slides, were deparaffinized and rehydrated prior to commencing. The slides were washed (5 min) in PBS (1.0 mM, pH 7.4), then transferred into equilibration buffer. After 10 min at room temperature in a humidified chamber, working strength TdT enzyme was added to positive sections while negative control sections, received PBS. Plastic coverslips were added and slides were placed in humidified chamber at 37 8C for 1 h. The slides were then washed with stop/wash buffer at 37 8C for 30 min, agitating after 10 min followed by a 5 min PBS wash. The slides were then blocked for endogenous peroxidase activity by immersion in 3% hydrogen peroxide. After two washes in PBS, the slides were placed in CAS blocking solution (Dako) for 30 min at room temperature. Excess blocking solution was tapped off and the antidigoxigenin conjugate was applied to all sections. Plastic coverslips were placed on top of sections and incubated for 30 min at room temperature in a humidified chamber, followed by two PBS washes (5 min each). Apoptotic cells were visualized with the addition of 0.05% (w:v) DAB chromagen (Sigma) for w1–2 min. The reaction was stopped by placing the slides into distilled H2O. The sections were lightly counterstained using neat Harris haematoxylin, dehydrated through a graded series of alcohols and mounted under glass with DPX (Sigma). TUNEL was applied to three to four randomly selected slides/whole cultured ovary, in order to reveal the extent of Reproduction (2008) 136 799–809

808

D Rosairo and others

apoptosis. On each slide, one section was used for TUNEL analysis, while the other acted as a negative control. Images were captured with a !20 objective and TdT-positively stained follicles were counted and categorized according to follicle (primordial, primary and preantral) class using analySIS Professional Imaging software, version 5.0. Follicles that stained positively for TdT enzyme were categorized as ‘atretic’ and represented as a percentage of the total number of follicles.

Statistical analysis Statistical significance was determined using SPSS software for Windows, version 14.0 (SPSS GmbH Software, Munich, Germany) by ANOVA in conjunction with a post hoc multiple comparison test (either Newman Keul’s: PCR data or Tukey’s: follicle counts). Each experiment was repeated at least three times, with n equal to a minimum of 3 per age group or treatment, in each experiment. The data are presented as the meanGS.D. P values of !0.05 compared with the appropriate control were regarded as statistically significant.

Declaration of interest There is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding This work was supported by the National Health and Medical Research Council of Australia (Regkeys 241000 and 198705).

Acknowledgements The authors would like to thank Dr Sarah Meachem and Saleela Runwanpura for their assistance with the TUNEL procedure. Sue Panckridge is thanked for her assistance in the preparation and submission of the manuscript. The financial support of the National Health and Medical Research Council of Australia (Regkeys 241000 and 198705) is acknowledged.

References Brunet A, Bonni A, Zigmond MJ, Lin MZ, Juo P, Hu LS, Anderson MJ, Arden KC, Blenis J & Greenberg ME 1999 Akt promotes cell survival by phosphorylating and inhibiting a Forkhead transcription factor. Cell 96 857–868. Chegini N & Flanders KC 1992 Presence of transforming growth factor-beta and their selective cellular localization in human ovarian tissue of various reproductive stages. Endocrinology 130 1707–1715. Christopher B 2000 Immunolocalization of transforming growth factorbeta1 during follicular development and atresia in the mouse ovary. Endocrine Journal 47 475–480. Derynck R, Zhang Y & Feng XH 1998 Smads: transcriptional activators of TGF-b responses. Cell 95 737–740. Drummond AE, Baillie AJ & Findlay JK 1999 Ovarian estrogen receptor alpha and beta mRNA expression: impact of development and estrogen. Molecular and Cellular Endocrinology 149 153–161. Drummond AE, Dyson M, Thean E, Groome NP, Robertson DM & Findlay JK 2000 Temporal and hormonal regulation of inhibin protein and subunit mRNA expression by post-natal and immature rat ovaries. Journal of Endocrinology 166 339–354. Reproduction (2008) 136 799–809

Drummond AE, Le MT, Ethier JF, Dyson M & Findlay JK 2002 Expression and localization of activin receptors, Smads, and beta glycan to the postnatal rat ovary. Endocrinology 143 1423–1433. Dunker N & Krieglstein K 2000 Targeted mutations of transforming growth factor-beta genes reveal important roles in mouse development and adult homeostasis. European Journal of Biochemistry 267 6982–6988. Ghiglieri C, Khatchadourian C, Tabone E, Hendrick JC, Benahmed M & Menezo Y 1995 Immunolocalization of transforming growth factor-b 1 and transforming growth factor-beta 2 in the mouse ovary during gonadotrophin-induced follicular maturation. Human Reproduction 10 2115–2119. Goddard I, Hendrick JC, Benahmed M & Morera AM 1995 Transforming growth factor beta receptor expression in cultured porcine granulosa cells. Molecular and Cellular Endocrinology 115 207–213. Gueripel X, Benahmed M & Gougeon A 2004 Sequential gonadotropin treatment of immature mice leads to amplification of transforming growth factor beta action, via upregulation of receptor-type 1, Smad 2 and 4, and downregulation of Smad 6. Biology of Reproduction 70 640–648. Hernandez ER, Hurwitz A, Payne DW, Dharmarajan AM, Purchio AF & Adashi EY 1990 Transforming growth factor-beta 1 inhibits ovarian androgen production: gene expression, cellular localization, mechanisms(s), and site(s) of action. Endocrinology 127 2804–2811. Ingman WV, Robker RL, Woittiez K & Robertson SA 2006 Null mutation in transforming growth factor beta1 disrupts ovarian function and causes oocyte incompetence and early embryo arrest. Endocrinology 147 835–845. Inoue K, Nakamura K, Abe K, Hirakawa T, Tsuchiya M, Oomori Y, Matsuda H, Miyamoto K & Minegishi T 2003 Mechanisms of action of transforming growth factor beta on the expression of follicle-stimulating hormone receptor messenger ribonucleic acid levels in rat granulosa cells. Biology of Reproduction 69 1238–1244. Juneja SC, Chegini N, Williams RS & Ksander GA 1996 Ovarian intrabursal administration of transforming growth factor beta 1 inhibits follicle rupture in gonadotropin-primed mice. Biology of Reproduction 55 1444–1451. Knecht M, Feng P & Catt K 1987 Bifunctional role of transforming growth factor-beta during granulosa cell development. Endocrinology 120 1243–1249. Knight PG & Glister C 2006 TGF-beta superfamily members and ovarian follicle development. Reproduction 132 191–206. Levacher C, Gautier C, Saez JM & Habert R 1996 Immunohistochemical localization of transforming growth factor beta 1 and beta 2 in the fetal and neonatal rat ovary. Differentiation 61 45–51. Liu X, Andoh K, Abe Y, Kobayashi J, Yamada K, Mizunuma H & Ibuki Y 1999 A comparative study on transforming growth factor-beta and activin A for preantral follicles from adult, immature, and diethylstilbestrolprimed immature mice. Endocrinology 140 2480–2485. May JV, Stephenson LA, Turzcynski CJ, Fong HW, Mau YH & Davis JS 1996 Transforming growth factor beta expression in the porcine ovary: evidence that theca cells are the major secretory source during antral follicle development. Biology of Reproduction 54 485–496. Mulheron GW & Schomberg DW 1990 Rat granulosa cells express transforming growth factor-beta type 2 messenger ribonucleic acid which is regulatable by follicle-stimulating hormone in vitro. Endocrinology 126 1777–1779. Mulheron GW, Danielpour D & Schomberg DW 1991 Rat thecal/interstitial cells express transforming growth factor-beta type 1 and 2, but only type 2 is regulated by gonadotropin in vitro. Endocrinology 129 368–374. Nilsson E, Doraiswamy V, Parrott JA & Skinner MK 2001 Expression and action of transforming growth factor beta (TGFb1, TGFb2, TGFb3) in normal bovine ovarian surface epithelium and implications for human ovarian cancer. Molecular and Cellular Endocrinology 182 145–155. Nilsson EE, Doraiswamy V & Skinner MK 2003 Transforming growth factorbeta isoform expression during bovine ovarian antral follicle development. Molecular Reproduction and Development 66 237–246. Ogawa T, Yogo K, Ishida N & Takeya T 2003 Synergistic effects of activin and FSH on hyperphosphorylation of Rb and G1/S transition in rat primary granulosa cells. Molecular and Cellular Endocrinology 210 31–38. www.reproduction-online.org

TGFB action in the ovary Park Y, Maizels ET, Feiger ZJ, Alam H, Peters CA, Woodruff TK, Unterman TG, Lee EJ, Jameson JL & Hunzicker-Dunn M 2005 Induction of cyclin D2 in rat granulosa cells requires FSH-dependent relief from FOXO1 repression coupled with positive signals from Smad. Journal of Biological Chemistry 280 9135–9148. Proetzel G, Pawlowski SA, Wiles MV, Yin M, Boivin GP, Howles PN, Ding J, Ferguson MW & Doetschman T 1995 Transforming growth factor-beta 3 is required for secondary palate fusion. Nature Genetics 11 409–414. Richards JS, Sharma SC, Falender AE & Lo YH 2002 Expression of FKHR, FKHRL1, and AFX genes in the rodent ovary: evidence for regulation by IGF-I, estrogen, and the gonadotropins. Molecular Endocrinology 16 580–599. Roy SK & Hughes J 1994 Ontogeny of granulosa cells in the ovary: lineagespecific expression of transforming growth factor beta 2 and transforming growth factor beta 1. Biology of Reproduction 51 821–830. Roy SK & Kole AR 1998 Ovarian transforming growth factor-beta (TGF-b) receptors: in-vitro effects of follicle stimulating hormone, epidermal growth factor and TGF-beta on receptor expression in human preantral follicles. Molecular Human Reproduction 4 207–214. Roy SK, Ogren C, Roy C & Lu B 1992 Cell-type-specific localization of transforming growth factor-beta 2 and transforming growth factor-beta 1 in the hamster ovary: differential regulation by follicle-stimulating hormone and luteinizing hormone. Biology of Reproduction 46 595–606. Sanford LP, Ormsby I, Gittenberger-de Groot AC, Sariola H, Friedman R, Boivin GP, Cardell EL & Doetschman T 1997 TGFb2 knockout mice have multiple developmental defects that are non-overlapping with other TGFbeta knockout phenotypes. Development 124 2659–2670.

www.reproduction-online.org

809

Schmid P, Cox D, van der Putten H, McMaster GK & Bilbe G 1994 Expression of TGF-b s and TGF-b type II receptor mRNAs in mouse folliculogenesis: stored maternal TGF-b 2 message in oocytes. Biochemical and Biophysical Research Communications 201 649–656. Shull MM, Ormsby I, Kier AB, Pawlowski S, Diebold RJ, Yin M, Allen R, Sidman C, Proetzel G, Calvin D et al. 1992 Targeted disruption of the mouse transforming growth factor-beta 1 gene results in multifocal inflammatory disease. Nature 359 693–699. Song L, Porter DG & Coomber BL 1999 Production of gelatinases and tissue inhibitors of matrix metalloproteinases by equine ovarian stromal cells in vitro. Biology of Reproduction 60 1–7. Teerds KJ & Dorrington JH 1992 Immunohistochemical localization of transforming growth factor-beta 1 and -beta 2 during follicular development in the adult rat ovary. Molecular and Cellular Endocrinology 84 R7–R13. Tsuchida K, Lewis KA, Mathews LS & Vale WW 1993 Molecular characterization of rat transforming growth factor-beta type II receptor. Biochemical and Biophysical Research Communications 191 790–795. Xiao S, Robertson DM & Findlay JK 1992 Effects of activin and folliclestimulating hormone (FSH)-suppressing protein/follistatin on FSH receptors and differentiation of cultured rat granulosa cells. Endocrinology 131 1009–1016.

Received 22 July 2008 First decision 9 September 2008 Accepted 9 September 2008

Reproduction (2008) 136 799–809