Implications
Angiogenin is a member of the ribonuclease A superfamily of proteins implicated in inducing angiogenesis. The present study provides new evidence for angiogenin-dependent regulation of follicle development. In particular, results indicate that angiogenin inhibits steroidogenesis of undifferentiated (small follicles) granulosa cells and stimulates mitogenesis of differentiated (large follicle) granulosa cells. The steroidogenic effects of angiogenin on large follicle theca cells were also inhibitory. Further studies will be required to elucidate cell-specific effects of angiogenin on ovarian cell proliferation and steroidogenesis during follicular development.
Introduction
Angiogenesis or the formation of blood vessels from pre-existing blood vessels is controlled by many different angiogenic factors (Lee et al., Reference Lee, Lee, Kang, Jeong and Chang1999; Ucuzian et al., Reference Ucuzian, Gassman, East and Greisler2010). Angiogenesis is involved in a variety of physiological processes such as wound repair and embryological development (Redmer and Reynolds, Reference Redmer and Reynolds1996; Przybylski, Reference Przybylski2009; Appelmann et al., Reference Appelmann, Liersch, Kessler, Mesters and Berdel2010), and is under the influence of anti-angiogenic and pro-angiogenic molecular mediators. One of the pro-angiogenic factors is angiogenin (ANG), a member of the ribonuclease A superfamily of 10 to 28 kDa proteins ascribed as a group of enzymes that have inherent substrate specificity and differing functional capacities (for review see: Tello-Montoliu et al., Reference Tello-Montoliu, Patel and Lip2006; Gupta et al., Reference Gupta, Haigh, Griffin and Wheeler2013; Premzl, Reference Premzl2014). However, unlike many of the RNase A superfamily members, ANG has been shown to have its own specific membrane receptor that mediates its effects (Hatzi and Badet, Reference Hatzi and Badet1999; Xu et al., Reference Xu, Monti and Hu2001). Angiogenin was first isolated from medium conditioned by human colon carcinoma (HT-29) cells (Fett et al., Reference Fett, Strydom, Lobb, Alderman, Bethune, Riordan and Vallee1985) and later found in normal human serum (Shapiro et al., Reference Shapiro, Strydom, Olson and Vallee1987), bovine milk (Maes et al., Reference Maes, Damart, Rommens, Montreuil, Spik and Tartar1988) and other mammalian serum (Bond et al., Reference Bond, Strydom and Vallee1993). Specific angiogenic effects of ANG are multifaceted (Gao and Xu, Reference Gao and Xu2008) and includes induction of endothelial cell proliferation (Hu et al., Reference Hu, Riordan and Vallee1997; Tsuji et al., Reference Tsuji, Sun, Kishimoto, Olson, Liu, Hirukawa and Hu2005), tumor cell adhesion (Soncin et al., Reference Soncin, Shapiro and Fett1994), and tumor cell invasion (Gho et al., Reference Gho, Yoon and Chae2002).
In domestic animals, ovarian vasculature surrounding follicles increases during normal follicular development (Yamada et al., Reference Yamada, Abe, Takehana, Hiraga, Iwasa and Hiratsuka1995; Jiang et al., Reference Jiang, Macchiarelli, Miyabayashi and Sato2002, Reference Jiang, Macchiarelli, Tsang and Sato2003; Moonmanee et al., Reference Moonmanee, Navanukraw, Uriyapongson, Kraisoon, Aiumlamai, Guntaprom, Rittirod, Borowicz and Redmer2013), but the molecular mechanisms that initiate ovarian angiogenesis are not well defined. In the bovine and human ovary, using in situ hybridization, ANG messenger RNA (mRNA) was localized in granulosa cells and oocytes (but not theca cells) of secondary and tertiary follicles, luteal cells of developing corpora lutea, and vascular endothelial and smooth muscle cells (Lee et al., Reference Lee, Lee, Kang, Jeong and Chang1999; Koga et al., Reference Koga, Osuga, Tsutsumi, Momoeda, Suenaga, Kuga, Fujiwara, Takai, Yano and Taketani2000). Malamitsi-Puchner et al. (Reference Malamitsi-Puchner, Sarandakou, Baka, Tziotis, Rizos, Hassiakos and Creatsas2001) reported that follicular fluid levels of ANG were significantly greater in human follicles yielding mature v. immature oocytes. Moreover, follicular fluid concentrations of ANG (ranging from 100 to 800 ng/ml) and progesterone were positively correlated in women (Koga et al., Reference Koga, Osuga, Tsutsumi, Momoeda, Suenaga, Kuga, Fujiwara, Takai, Yano and Taketani2000). Therefore, we hypothesized that ANG, in addition to its angiogenic effects, would stimulate steroidogenesis and increase proliferation of both granulosa and theca cells. Therefore, the objectives of this study were to determine the effect of ANG on granulosa and theca cell proliferation and steroid production in vitro.
Materials and methods
Hormones and reagents
The hormones and reagents used in cell culture were: ovine FSH (NIDDK-oFSH-20; activity: 175 X NIH-FSH-S1 U/mg) and ovine LH (NIDDK-oLH-26; activity: 1.0 X NIH-LH-S1 U/mg) from the National Hormone & Pituitary Program (Torrance, CA, USA), recombinant human ANG and IGF-1 (IGF1) from R&D Systems (Minneapolis, MN, USA), testosterone from Steraloids (Wilton, NH, USA), fetal calf serum (FCS) from EquiTech-Bio Inc. (Kerrville, TX, USA), and collagenase and DNase from Sigma-Aldrich Corp. (St. Louis, MO, USA). Human ANG and bovine ANG proteins share 66% amino acid homology (http://blast.ncbi.nlm.nih.gov/Blast.cgi).
Cell culture
Ovaries from non-pregnant beef cows were collected from a local slaughterhouse, and based on surface diameter, granulosa cells were collected from small (1 to 5 mm) and large (8 to 22 mm) follicles as previously described (Langhout et al., Reference Langhout, Spicer and Geisert1991; Stewart et al., Reference Stewart, Spicer, Hamilton and Keefer1995; Spicer et al., Reference Spicer, Aad, Allen, Mazerbourg, Payne and Hsueh2008). Small and large follicles were bisected after aspiration of follicular fluid and granulosa cells separated from the theca interna via blunt dissection, and the theca interna was torn into small pieces, rinsed with basal serum-free medium (1 : 1 Dulbecco’s modified Eagle’s medium and Ham’s F-12; 0.12 mM gentamycin, 2.0 mM glutamine, and 38.5 mM sodium bicarbonate; all obtained from Sigma-Aldrich Corp.) and enzymatically digested for 1 h at 37°C as previously described (Stewart et al., Reference Stewart, Spicer, Hamilton and Keefer1995; Spicer and Chamberlain, Reference Spicer and Chamberlain1998; Spicer et al., Reference Spicer, Aad, Allen, Mazerbourg, Payne and Hsueh2008). Briefly, after digestion, the non-digested tissue was filtered out, and theca cells were then centrifuged at 50×g for 5 min, supernatant discarded and pellet washed with serum-free medium. The purity of theca cells prepared this way was >90% (Spicer and Stewart, Reference Spicer and Stewart1996; Spicer et al., Reference Spicer, Aad, Allen, Mazerbourg, Payne and Hsueh2008). The theca and granulosa cells were re-suspended in serum-free media containing collagenase and DNase at 1.25 and 0.5 mg/ml, respectively, to prevent cell clumping before plating.
Viable cells (2×105 in 25 to 45 μl of medium/well) were plated in each well of 24-well Falcon multiwell plates (Becton Dickinson, Lincoln Park, NJ, USA) containing 1 ml of basal medium with 10% FCS (v/v). Cells were cultured at 38.5°C in 10% FCS (v/v) for the first 48 h with a medium change at 24 h. Cells were then washed twice with serum-free medium and the various treatments (see below) applied in serum-free medium (1 ml per well) for 48 h with a medium change at 24 h. After 48 h, medium was collected for steroid quantification via radioimmunoassays (RIA) and cells were collected for cell enumeration via Coulter counting (see below). The concentrations of LH, FSH and IGF1 were selected based on previous studies (Spicer and Stewart, Reference Spicer and Stewart1996; Spicer and Chamberlain, Reference Spicer and Chamberlain1998; Spicer et al., Reference Spicer, Chamberlain and Maciel2002). Because steroid production in this culture system is maximized with combined gonadotropins and growth factor treatment, only weakly responsive to gonadotropins alone, and not responsive to growth factors alone (Stewart et al., Reference Stewart, Spicer, Hamilton and Keefer1995; Spicer and Chamberlain, Reference Spicer and Chamberlain1998; Spicer et al., Reference Spicer, Chamberlain and Maciel2002), gonadotropins were used in combination with IGF1 for most experiments. The concentrations of ANG used for experiments were selected based on studies indicating that concentrations of ANG in bovine plasma (Bond and Vallee, Reference Bond and Vallee1988; Chang et al., Reference Chang, Jeong, Park, Ahn, Choi, Chae, Namgoong and Chung1997) and human follicular fluid (Koga et al., Reference Koga, Osuga, Tsutsumi, Momoeda, Suenaga, Kuga, Fujiwara, Takai, Yano and Taketani2000; Malamitsi-Puchner et al., Reference Malamitsi-Puchner, Sarandakou, Baka, Tziotis, Rizos, Hassiakos and Creatsas2001; Kawano et al., Reference Kawano, Zeineh Hasan, Fukuda, Mine and Miyakawa2003) range between 0.4 and 800 ng/ml.
Experimental design
Experiment 1 was designed to determine the dose-response effects of ANG on steroidogenesis and number of granulosa cells of small follicles. Granulosa cells from small follicles were selected for this experiment because we have shown that they grow and respond well to growth factor treatments (Spicer et al., Reference Spicer, Chamberlain and Maciel2002). Granulosa cells from small follicles were cultured for 48 h in the presence of 10% FCS, and then cells were washed and incubated in serum-free medium in the presence of 0, 30 or 100 ng/ml of ANG without or with 30 ng/ml of FSH and/or 30 ng/ml of IGF1 for 48 h. Medium was changed every 24 h and contained 500 ng/ml of testosterone as an estrogen precursor (Spicer and Chamberlain, Reference Spicer and Chamberlain1998). Cells were counted and medium samples were collected for progesterone (P4) and estradiol (E2) determinations via RIA (see below).
Experiment 2 was designed to determine the dose-response effects of ANG on steroidogenesis and number of theca cells of large follicles. Theca cells from large follicles were selected for this experiment because we have shown that they grow and respond well to growth factor treatments (Spicer et al., Reference Spicer, Aad, Allen, Mazerbourg, Payne and Hsueh2008). Theca cells from large follicles were cultured for 48 h in the presence of 10% FCS, and then cells were washed and incubated in serum-free medium in the presence of 0, 30 or 100 ng/ml of ANG without or with 30 ng/ml of LH and/or 30 ng/ml of IGF1 for 48 h. Medium was changed every 24 h. Cells were counted and medium samples were collected for P4 and androstenedione determinations via RIA (see below).
Experiment 3 was designed to compare the effects of high dose (i.e. 300 ng/ml) ANG on steroidogenesis and number of granulosa cells isolated from small and large follicles. Because experiment 1 showed minor effects of 30 and 100 ng/ml of ANG on small follicle granulosa cell numbers or steroid production, studies were conducted using 300 ng/ml of ANG with both small and large follicle granulosa cells. Cells from small and large follicles were cultured for 48 h in the presence of 10% FCS, and then cells were washed and incubated in serum-free medium in the presence of 0 or 300 ng/ml of ANG without or with 30 ng/ml of FSH and/or 30 ng/ml of IGF1 for 48 h. Medium was changed every 24 h and contained 500 ng/ml of testosterone as an estrogen precursor (Spicer and Chamberlain, Reference Spicer and Chamberlain1998). Cells were counted and medium samples were collected for P4 and E2 determinations via RIA (see below).
Experiment 4 was designed to compare the effects of high dose (i.e. 300 ng/ml) ANG on steroidogenesis and number of theca cells isolated from small and large follicles. Because experiment 2 showed minor effects of 30 and 100 ng/ml of ANG on large follicle theca cell numbers or steroid production, studies were conducted using 300 ng/ml of ANG with both small and large follicle theca cells. For this experiment, small follicle theca cells were obtained from follicles 3 to 6 mm in diameter as previously described (Spicer et al., Reference Spicer, Aad, Allen, Mazerbourg, Payne and Hsueh2008). Cells from small and large follicles were cultured for 48 h in the presence of 10% FCS, and then cells were washed and incubated in serum-free medium in the presence of 0 or 300 ng/ml of ANG without or with 30 ng/ml of LH and/or 30 ng/ml of IGF1 for 48 h. Medium was changed every 24 h. Cells were counted and medium samples were collected for P4 and androstenedione determinations via RIA (see below).
Experiment 5 was designed to test the effect of ANG on large follicle granulosa cell numbers. Cells were cultured for 48 h in 10% FCS medium, and then treated for an additional 24 h or 48 h with either 0 or 300 ng/ml of ANG in the presence of 10% FCS. Separate cultures were terminated at 0, 24 and 48 h post-treatment, and cells were counted. Dose of ANG was selected based on results from experiments 1 and 3.
Experiment 6 was designed to determine the effect of ANG on large follicle granulosa and theca cell proliferation as measured by 3H-thymidine incorporation into DNA. Because experiments 1 to 4 showed an effect of ANG on cell numbers only in the presence of IGF1, this experiment evaluated the effect of ANG in the presence of IGF1. After 48 h in 10% FCS, cells were serum-starved for 24 h by culturing in serum-free medium, medium changed, and then cells cultured for an additional 40 h in serum-free medium with either 0 or 300 ng/ml of angiogenin in the presence of 1 μCi of 3H-thymidine and 30 ng/ml of IGF1 to assess DNA synthesis as previously described (Spicer and Aad, Reference Spicer and Aad2007; Spicer et al., Reference Spicer, Aad, Allen, Mazerbourg, Payne and Hsueh2008).
Experiment 7 was conducted to determine the effect ANG on steroidogenic gene expression in granulosa and theca cells. Granulosa cells from small follicles and theca cells from large follicles were cultured for 48 h in 10% FCS, medium changed, and then cells cultured for an additional 24 h in serum-free medium with either no treatment or 300 ng/ml of ANG. Granulosa cells were concomitantly treated with 30 ng/ml of FSH and IGF1 and theca cells were treated concomitantly with 30 ng/ml of LH and IGF1. Doses of FSH, LH and IGF1 were selected based on previous studies (Spicer et al., Reference Spicer, Chamberlain and Maciel2002; Lagaly et al., Reference Lagaly, Aad, Grado-Ahuir, Hulsey and Spicer2008). Gonadotropins were added to all treatments because IGF1 alone has little or no effect on steroid production (Lagaly et al., Reference Lagaly, Aad, Grado-Ahuir, Hulsey and Spicer2008; Spicer et al., Reference Spicer, Aad, Allen, Mazerbourg, Payne and Hsueh2008). After 24 h of treatment, cells were lysed in 0.5 ml of TRIzol for RNA extraction (see below) and quantification of CYP11A1, CYP19A1 or CYP17A1 mRNA.
RNA extraction and quantification
Total RNA was extracted using TRIzol reagent protocol (Life Technologies, Carlsbad, CA, USA), and RNA was quantitated by spectrophotometry at 260 nm using a NanoDrop ND-1000 (NanoDrop Technologies, Wilmington, DE, USA) as previously described (Voge et al., Reference Voge, Santiago, Aad, Goad, Malayer and Spicer2004; Aad et al., Reference Aad, Voge, Santiago, Malayer and Spicer2006; Schreiber and Spicer, Reference Schreiber and Spicer2012). Quantification of gene expression was conducted by fluorescent one-step real-time PCR as previously described (Spicer and Aad, Reference Spicer and Aad2007; Grado-Ahuir et al., Reference Grado-Ahuir, Aad and Spicer2011). The bovine CYP11A1, CYP17A1 and CYP19A1 primer and probe sequences and information are described by Lagaly et al. (Reference Lagaly, Aad, Grado-Ahuir, Hulsey and Spicer2008). The internal standard was 18S ribosomal RNA. Data analysis was done using the comparative threshold cycle (C
t
) method as previously described (Voge et al., Reference Voge, Santiago, Aad, Goad, Malayer and Spicer2004; Aad et al., Reference Aad, Voge, Santiago, Malayer and Spicer2006). Briefly, the ∆C
t
was determined by subtracting the 18S C
t
value from the target gene unknown value. For each target gene and within each experiment, the ∆∆C
t
was determined by subtracting the higher ∆C
t
(the least expressed unknown) from all other ∆C
t
values. Fold changes in target gene mRNA abundance were calculated as being equal to
$${\rm 2}^{{{\minus}\Delta \Delta C_{t} }} $$
.
Determination of steroid concentrations and cell numbers
Medium was collected from individual wells and frozen at –20°C for subsequent determination of concentrations of P4, E2 and/or androstenedione via RIA as previously described (Langhout et al., Reference Langhout, Spicer and Geisert1991; Stewart et al., Reference Stewart, Spicer, Hamilton and Keefer1995; Spicer and Chamberlain, Reference Spicer and Chamberlain1998). The intra- and interassay coefficients of variation were 5.6% and 11% for the progesterone RIA, 6.5% and 12% for the estradiol RIA, and 4.1% and 7.3% for the androstenedione RIA.
Numbers of cells were determined via Coulter counting as previously described (Langhout et al., Reference Langhout, Spicer and Geisert1991; Stewart et al., Reference Stewart, Spicer, Hamilton and Keefer1995; Spicer et al., Reference Spicer, Aad, Allen, Mazerbourg, Payne and Hsueh2008) and used to calculate steroid production on a ng or pg per 105 cell basis. Briefly, cells were gently washed twice with 0.9% saline (500 µl; w/v), exposed to 500 µl of trypsin solution (0.25%, w/v) for 20 min at 25°C, scraped from each well, and counted using a Coulter counter (model Z2; Beckman Coulter Inc., Miami, FL, USA).
Cell aggregates were disrupted via pipetting the cell suspension back and forth through a 500 μl pipette tip three to five times, and diluted in 9 ml of 0.9% saline (w/v) before counting.
Statistical analysis
Data are presented as the least squares means (±SE) of measurements from three or more individual pools of large and small follicle granulosa/theca cells used as experimental replicates. Each biological replicate (i.e. pool of cells) was conducted on independent pools of cells. Each pool of small follicle granulosa cells were collected from 10 to 30 ovaries (n=5 to 15 cattle) yielding 6 to 8 ml of follicular fluid. Each of the large follicle granulosa/theca cell pools was obtained from 7 to 10 follicles from at least five animals. Small follicle theca cells were obtained from 6 to 20 ovaries (n=3 to 10 animals). Within each replicated experiment, treatments were applied to each pool of cells in duplicate or triplicate culture wells. Steroid production was expressed as ng or pg/105 cells per 24 h, and cell numbers at the termination of each experiment were used for this calculation. Specific differences in cell numbers and steroid production among treatments were determined via ANOVA using GLM procedure of SAS (Statistical Analysis System, Cary, NC, USA) and Fisher’s protected least significant difference procedure (Ott, Reference Ott1977). Significance was declared at P<0.05.
Results
Experiment 1: dose response of ANG on cell numbers and steroidogenesis of small follicle granulosa cells
Treatment of granulosa cells with IGF1 alone increased (P<0.05) cell numbers by 54% to 73% (Table 1), however none of the doses of ANG (i.e. 30 or 100 ng/ml) affected (P>0.10) control or IGF1-induced granulosa cell numbers (Table 1). Alone FSH had no effect (P>0.05) on cell numbers but FSH significantly enhanced the IGF1-induced increase (P<0.001) in cell numbers (Table 1). Dose of ANG had no significant effect on E2 production (Figure 1a). FSH and IGF1 synergized to stimulate (P<0.01) E2 production by 6.6-fold, and ANG had no significant effect on this FSH plus IGF1-induced E2 production (Figure 1a); alone neither FSH nor IGF1 affected (P>0.10) E2 production. Both IGF1 and FSH increased P4 production and 100 ng/ml of ANG reduced (P<0.05) the FSH plus IGF1-induced P4 production by 16% (Figure 1b).
Effect of angiogenin on basal, FSH- and IGF1-induced estradiol (a) and progesterone (b) production by small-follicle granulosa cells (experiment 1). Cells were cultured for 48 h as described in ‘Materials and methods’ section, and then treated for an additional 48 h with 0, 30 or 100 ng/ml of angiogenin (ANG) and: Control (no IGF1 or FSH), IGF1 (30 ng/ml), FSH (30 ng/ml) or FSH plus IGF1. Values are means±SEM of three separate experiments (n=6). *Within a panel and treatment, mean differs (P<0.05) from its respective 0 ANG mean. #Within a panel, mean differs (P<0.05) from its respective Control mean. +Within a panel, mean differs (P<0.05) from its respective IGF1 alone or FSH alone treatment mean.
Effect of recombinant human angiogenin on granulosa and theca cell numbers in experiments 1 and 2
Granulosa cells (GC) from small (1 to 5 mm; experiment 1) and theca cells (TC) from large (8 to 22 mm; experiment 2) bovine follicles were cultured as described in ‘Materials and methods’ section, and treated for 48 h with FSH (for GC), LH (for TC) and/or IGF1 (0 or 30 ng/ml) and 0, 30 or 100 ng/ml of angiogenin.
a,b,cWithin an experiment, means without a common letter differ (P<0.05).
1 SEM for n=6.
Experiment 2: dose response of ANG on steroidogenesis of large follicle theca cells
Treatment of theca cells with IGF1 increased (P<0.05) cell numbers by 31% to 56% (Table 1). Neither LH nor ANG (30 or 100 ng/ml) affected (P>0.10) theca cell numbers induced by IGF1 (Table 1). LH increased (P<0.05) both androstenedione and P4 production, and ANG (30 and 100 ng/ml) decreased (P<0.05) IGF1 plus LH-induced androstenedione production by 22% to 23%, but ANG had no effect (P>0.10) on basal, LH-induced or IGF1-induced androstenedione production (Figure 2a). None of the doses of ANG affected (P>0.10) P4 production by theca cells (Figure 2b).
Effect of angiogenin on basal, LH- and IGF1-induced androstenedione (a) and progesterone (b) production by large follicle theca cells (experiment 2). Cells were cultured for 48 h as described in ‘Materials and methods’ section, and then treated for an additional 48 h with 0, 30 or 100 ng/ml of angiogenin (ANG) and: Control (no IGF1 or LH), IGF1 (30 ng/ml), LH (30 ng/ml) or LH plus IGF1. Values are means±SEM of three separate experiments (n=6). *Within a panel and treatment, mean differs (P<0.05) from its respective 0 ANG mean. #Within a panel, mean differs (P<0.05) from its respective Control mean. +Within a panel, mean differs (P<0.05) from its respective LH alone treatment mean.
Experiment 3: effect of high-dose of ANG on steroidogenesis of small and large follicle granulosa cells
Angiogenin had no effect (P>0.10) on numbers of small follicle granulosa cells, but increased (P<0.05) cell numbers in FSH+IGF1-treated large follicle granulosa cells (Table 2). Treatment with FSH and IGF1 increased (P<0.05) cell numbers by 2.3-fold and 4.8-fold above controls in small and large follicle granulosa cell cultures, respectively. Angiogenin decreased (P<0.05) P4 production by 12% in small follicle granulosa cells treated with FSH+IGF1 and by 30% in control (untreated) large follicle granulosa cells (Table 2). In small follicle granulosa cells, FSH and IGF1 stimulated (P<0.01) E2 production by 5.7-fold, and ANG decreased this hormone-induced E2 production by 27% (Table 2). In large follicle granulosa cells, ANG had no effect (P>0.10) on E2 production (Table 2). To compare across granulosa cell experiments, data (basal and FSH plus IGF1 treatment) from experiments 1 and 3 were expressed as fold of controls and shown in Supplementary Figure S1.
Effect of 2-day treatment of 300 ng/ml of angiogenin (ANG) on basal and IGF1 plus FSH-induced progesterone production, estradiol production and numbers of granulosa cells from small (1 to 5 mm) and large (⩾8 mm) bovine follicles of experiment 3
a,b,cWithin a column, means without a common superscript differ (P<0.05).
1 SEM for n=6.
Experiment 4: effect of high-dose ANG on steroidogenesis of small and large follicle theca cells
LH decreased (P<0.05) small follicle theca cell numbers induced by IGF1 but had no effect (P>0.10) on numbers of large follicle theca cells (Table 3). Angiogenin (300 ng/ml) restored this decrease induced by LH (Table 3) in small follicle theca cells, and increased (P<0.05) numbers of IGF1-treated large follicle theca cells by 23% but had no effect on LH+IGF1-treated large follicle theca cells (Table 3). Angiogenin had no effect (P>0.10) on P4 production by small follicle theca cells (Table 3), but decreased (P<0.05) P4 production in IGF1-treated large follicle theca cells by 17% (Table 3). In both small and large follicle theca cells, ANG had no effect (P>0.10) on IGF1-induced or LH+IGF1-induced androstenedione production (Table 3). To compare across theca cell experiments, data (basal and LH plus IGF1 treatment) from experiments 2 and 4 were expressed as fold of controls and shown in Supplementary Figure S2.
Effect of 2-day treatment of 300 ng/ml of angiogenin (ANG) on IGF1- and IGF1 plus LH-induced progesterone production, androstenedione production and numbers of theca cells from small (3 to 6 mm) and large (⩾8 mm) bovine follicles of experiment 4
a,b,cWithin a column, means without a common superscript differ (P<0.05).
1 SEM for n=9.
Experiment 5: effect of ANG on granulosa and theca cell proliferation induced by 10% FCS
After 2 days of treatment, ANG (300 ng/ml) further enhanced (P<0.05) proliferation of large follicle granulosa cells stimulated by 10% FCS (Figure 3a). Granulosa cells grew 2.4- and 2.7-fold between day 0 and 2 in Control and ANG-treated cultures, respectively (P<0.05; Figure 3a). Cell numbers in ANG-treated cells did not differ (P>0.10) from controls on day 1 of treatment (Figure 3a).
Effect of angiogenin on proliferation of large follicle granulosa cells induced by 10% FCS (a) (experiment 5) and 3H-thymidine incorporation into large follicle granulosa and theca cells (b) treated with IGF1 (experiment 6). (a) Cells were cultured for 48 h as described in ‘Materials and methods’ section, and then treated for an additional 48 h with 0 or 300 ng/ml of angiogenin in the presence of 10% FCS; cells were enumerated via Coulter counting. (b) Cells were cultured for 48 h as described in ‘Materials and methods’ section, serum-starved for 24 h, and then treated for an additional 40 h with 3H-thymidine and IGF1 (30 ng/ml) and either 0 or 300 ng/ml of angiogenin. *Within a panel, asterisk (*) indicates mean differs (P<0.05) from its respective control (0 ANG) mean. Values are means±SEM of three separate experiments (n=6).
Experiment 6: effect of ANG on granulosa and theca cell proliferation induced by IGF1
Treatment of bovine granulosa cells from large follicles with 300 ng/ml of ANG increased (P<0.05) IGF1-induced 3H-thymidine incorporation into DNA by 2.6-fold (Figure 3b). Angiogenin (300 ng/ml) had no effect (P>0.10) on large follicle theca cell proliferation as measured by 3H-thymidine incorporation into DNA (Figure 3b).
Experiment 7: effect of ANG on steroidogenic gene expression in granulosa and theca cells
Treatment of 300 ng/ml of ANG to granulosa cells from small follicles and theca cells from large follicles had no effect (P>0.10) on CYP11A1 mRNA abundance (Figure 4). However, ANG at 300 ng/ml increased (P<0.05) CYP19A1 mRNA abundance in granulosa cells and had no effect (P>0.10) on CYP17A1 mRNA abundance in theca cells (Figure 4).
Effect of angiogenin (ANG) on steroidogenic enzyme gene expression in granulosa and theca cells (experiment 7). (a) Effect of effect of ANG (300 ng/ml) on CYP19A1 (aromatase) and CYP11A1 (side-chain cleavage enzyme) mRNA abundance in small follicle granulosa cells. (b) Effect of ANG (300 ng/ml) on CYP17A1 and CYP11A1 mRNA abundance in large follicle theca cells. Granulosa cells from small follicles and theca cells from large follicles were cultured for 48 h in the presence of 10% FCS, and then treated in serum-free medium with ANG for 24 h. Granulosa cells were also concomitantly treated with 30 ng/ml of IGF1 and 30 ng/ml of FSH, and theca cells were concomitantly treated with 30 ng/ml of IGF1 and 30 ng/ml of LH. Values are means of three separate experiments (±SEM) and normalized to constitutively expressed 18S ribosomal RNA. *Mean differs (P<0.05) from its respective control (0 ng/ml of ANG).
Discussion
Results of the present study revealed that ANG stimulated large follicle granulosa cell proliferation and inhibited small follicle granulosa cell P4 and E2 production and large follicle theca cell P4 and androstenedione production, whereas ANG had no effect on large follicle theca cell proliferation.
For the first time, the present experiments showed that presumed physiological concentrations of ANG (i.e. 30 to 300 ng/ml) were bioactive in both granulosa and theca cell cultures. Although not known for cattle, follicular fluid levels of ANG in women range from 0.4 to 800 ng/ml (Koga et al., Reference Koga, Osuga, Tsutsumi, Momoeda, Suenaga, Kuga, Fujiwara, Takai, Yano and Taketani2000; Malamitsi-Puchner et al., Reference Malamitsi-Puchner, Sarandakou, Baka, Tziotis, Rizos, Hassiakos and Creatsas2001; Kawano et al., Reference Kawano, Zeineh Hasan, Fukuda, Mine and Miyakawa2003). Within this concentration range, ANG inhibited granulosa and theca cell steroidogenesis but stimulated granulosa cell proliferation, suggesting a possible regulatory role in follicular development. We hypothesize that ANG, while stimulating blood vessel development, may also promote proliferation of large follicle granulosa cells, while slowing differentiation of small follicle granulosa cells. Presumably, this ANG-induced suppression of small follicle differentiation would allow small follicles to grow and develop its vasculature before they are able to respond to increased gonadotropins during the preovulatory period.
Angiogenin effects on numbers of granulosa or theca cells have not been reported previously and varied with dose, cell type and hormone treatment. At 30, 100 and 300 ng/ml, ANG had no significant effect on numbers of small follicle granulosa cells or numbers of large follicle theca cells. However, at 300 ng/ml, ANG increased proliferation of large follicle granulosa cells in the presence of IGF1 and 10% FCS. These results indicate a change in ANG response as follicles develop. A previous study indicated that ANG stimulates 3H-thymidine incorporation into human umbilical venous endothelial cells grown in medium containing 5% FCS and 5 ng/ml of basic fibroblast growth factor (Hu et al., Reference Hu, Riordan and Vallee1997; Liu et al., Reference Liu, Yu, Xu, Riordan and Hu2001). Cell survival and anti-apoptotic effects of ANG have been reported in other cell models (Li and Hu, Reference Li and Hu2012; Saikia et al., Reference Saikia, Jobava, Parisien, Putnam, Krokowski, Gao, Guan, Yuan, Jankowsky, Feng, Hu, Pusztai-Carey, Gorla, Sepuri, Pan and Hatzoglou2014). Thus, effects of ANG on follicular atresia should be investigated. Also, further research will be required to determine the mechanisms by which differentiated granulosa cells from large follicles are more responsive to the stimulatory effect of ANG on proliferation than less differentiated granulosa cells of small follicles, but likely involve changes in the IGF1 intracellular response system, and/or a change in numbers of ANG receptors.
In the present study, ANG had weak inhibitory effects on P4 production by small follicle granulosa cells concomitantly treated with FSH and IGF1, a condition in which cyclic adenosine monophosphate is dramatically elevated (Davoren et al., Reference Davoren, Hsueh and Li1985; Zhang et al., Reference Zhang, Garmey and Veldhuis2000). In untreated large follicle granulosa cells and in IGF1-treated large follicle theca cells, ANG also weakly inhibited P4 production. However, these small decreases in P4 production were not accompanied with any change in abundance of CYP11A1 mRNA suggesting that the mechanism for P4 inhibition by ANG is not mediated via a change in CYP11A1 expression. Nonetheless, the present studies indicate for the first time that ANG, in addition to its purported role in inducing angiogenesis, may regulate steroidogenesis in granulosa and theca cells. In small follicle granulosa cells, the paradoxical increase in CYP19A1 mRNA induced by ANG while E2 production remained unchanged at the same dose of ANG will require further elucidation. ANG did not affect E2 production by large follicle granulosa cells indicating that its effect on steroidogenesis in large follicle granulosa cells is exclusive to P4 production. In large follicle theca cells, 30 and 100 ng/ml of ANG decreased LH plus IGF1-induced androstenedione production but when tested at 300 ng/ml, ANG had no effect on androstenedione production or abundance of CYP17A1 mRNA, suggesting a bi-phasic response of theca cells to ANG. However, this inhibitory effect of ANG on androstenedione production was weak (i.e. 22% to 23% inhibition) and theca cell P4 production was only inhibited by 17% in large follicle theca cells treated with 300 ng/ml of ANG, suggesting a minor role for ANG in regulating theca cell steroidogenesis. Immunohistochemical staining and in situ hybridization in cattle showing that theca cells localize ANG protein but not ANG mRNA provides further support of the idea that ANG produced by granulosa cells communicates with theca cells (Lee et al., Reference Lee, Lee, Kang, Jeong and Chang1999). Further studies will be required to elucidate the developmental cell-specific effects of ANG on ovarian cell proliferation and steroidogenesis.
The mechanism of action and intracellular signaling pathway of ANG that affect granulosa and theca cell mitosis is unknown. In a recent study, ANG-induced proliferation of human glioblastoma U87MG cells was reported to act via a nuclear factor-κB pathway (Xia et al., Reference Xia, Fu, Cai, Wang, Chen, Xing, Wang, Zou, Xu and Xu2015). In another study, the protein kinase B/Akt pathway was reported to be induced by ANG in human umbilical vein endothelial cells (Kim et al., Reference Kim, Kang, Kim, Kang and Chang2007). Moreover, Saikia et al. (Reference Saikia, Jobava, Parisien, Putnam, Krokowski, Gao, Guan, Yuan, Jankowsky, Feng, Hu, Pusztai-Carey, Gorla, Sepuri, Pan and Hatzoglou2014) showed that ANG-generated transfer RNA halves (tiRNAs) and survival of mouse embryonic cortical neuronal cells involves a cytochrome c interaction that inhibits apoptosome formation and activity. Whether any of these pathways are sensitive to ANG in bovine granulosa and theca cells will require further study.
In summary, results of the present study provide new evidence for ANG-dependent regulation of proliferation and steroidogenesis in granulosa and thecal cells that may influence follicle development. In particular, results indicate that ANG inhibits steroidogenesis of undifferentiated (small follicles) granulosa cells and stimulates mitogenesis of differentiated (large follicle) granulosa cells. The steroidogenic effects of ANG on large follicle theca cells although weak were also inhibitory. We conclude that ANG may target granulosa and theca cells in cattle, stimulating proliferation and inhibiting steroidogenesis, but additional research is needed to understand the mechanism of action of ANG in granulosa and theca cells, as well as its precise role in folliculogenesis.
Acknowledgments
The authors thank the Oklahoma State University Wentz Project scholarship program (Lew Wentz Foundation) for financial support of A. Burress; the Bill & Melinda Gates Foundation for financial support of J. Dentis; Dr. A. F. Parlow, National Hormone & Pituitary Program, (Torrance, CA, USA) for purified LH and FSH; and Creekstone Farms (Arkansas City, KS, USA) for their generous donations of bovine ovaries. Approved for publication by the Director, Oklahoma Agric. Exp. Sta., and supported in part under project H-2510. This work was supported by project grant #R15-HD066302-01 from the National Institutes of Health.
Supplementary Material
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/S1751731116002044
