General and Comparative Endocrinology 176 (2012) 173–181

Contents lists available at SciVerse ScienceDirect

General and Comparative Endocrinology

journal homepage: www.elsevier .com/locate /ygcen

Basic fibroblast growth factor suppresses meiosis and promotes mitosisof ovarian germ cells in embryonic chickens

Bin He a,1, Jinxing Lin a,b,1, Jie Li a, Yuling Mi a, Weidong Zeng a, Caiqiao Zhang a,⇑a Department of Veterinary Medicine, College of Animal Sciences, Zhejiang University, Hangzhou 310058, Chinab Shanghai Laboratory Animal Research Center, Shanghai 201203, China

a r t i c l e i n f o a b s t r a c t

Article history:Received 12 November 2011Revised 12 January 2012Accepted 13 January 2012Available online 28 January 2012

Keywords:Basic fibroblast growth factorMeiosisMitosisGerm cellChickenOvary

0016-6480/$ - see front matter � 2012 Elsevier Inc. Adoi:10.1016/j.ygcen.2012.01.012

⇑ Corresponding author. Address: College of AnimalNo. 866 Yuhangtang Road, Hangzhou 310058, China.

E-mail address: cqzhang@zju.edu.cn (C. Zhang).1 These authors contributed equally to this work.

Basic fibroblast growth factor (bFGF or FGF2) plays diverse roles in regulating cell proliferation, migrationand differentiation during embryo development. In this study, the effect of bFGF on ovarian germ celldevelopment was investigated in the embryonic chicken by in vitro and in vivo experiments. Resultsshowed that a remarkable decrease in bFGF expression in the ovarian cortex was manifested during mei-osis progression. With ovary organ culture, we revealed that meiosis was initiated after retinoic acid (RA)treatment alone but was decreased after combined bFGF treatment that was detected by real time RT-PCR, fluorescence immunohistochemistry and Giemsa staining. Further, no significant difference inmRNA expression of either RA metabolism-related enzymes (Raldh2 and Cyp26b1) or RA receptorswas displayed after bFGF challenge. This result suggests that the suppression of bFGF on meiosis wasunlikely through inhibition of RA signaling. In addition, as a mitogen, bFGF administration increased germcell proliferation (via BrdU incorporation) in cultured organ or cells in vitro and also in developingembryos in vivo. In contrast, blockade of bFGF action by SU5402 (an FGFR1 antagonist) or inhibition ofprotein kinase C signaling showed inhibited effect of bFGF on mitosis. In conclusion, bFGF suppressesRA-induced entry of germ cells into meiosis to ensure embryonic ovarian germ cells to maintain at undif-ferentiated status and accelerate germ cell proliferation by binding with FGFR1 involving PKC activationin the chicken.

� 2012 Elsevier Inc. All rights reserved.

1. Introduction

Germ cells play a uniquely important role in biology becausethey are the only cell lineage to ensure the transmission of geneticinformation from one generation to the next. Two main events oc-cur during the early development of the germ cell lineage in theembryonic ovary: mitosis (cell proliferation) and the initiation ofmeiosis (cell differentiation). But how germ cells switch from mito-sis to meiosis represents a key question in reproductive biology.Numerous studies indicate that various hormones, growth factors,cytokines and microRNAs are involved in regulating this physiolog-ical process [5,6,16,17,18,23,24,40].

In the mouse, to boost the number, germ cells proliferate mitot-ically during migration to the genital ridges and also for 1–2 daysafter allocation and then undergo a cessation in cell division and en-ter meiosis in the ovary [29]. Recent studies demonstrate that thevitamin A derivative retinoic acid (RA) triggers germ cells to entermeiosis [1,5,23]. Moreover, the meiosis-inducing function of RA

ll rights reserved.

Sciences, Zhejiang University,Fax: +86 571 88982976.

has been proved to operate also in the chickens [37]. Nevertheless,in the female chicken embryo, the signal for meiosis is activatedasynchronously during germ cell population. In fact, while somecells may enter meiotic prophase, others are still undergoing mitosis[19,37]. It has been reported that the germ cell population as a wholeis still proliferating up to embryonic day 17 (E17) of incubation [19].Same phenomenon is reported in human that meiosis initiationappears asynchronous, while more and more germ cells initiatemeiosis, some oogonia that express pluripotent stem cell markerscontinue proliferating until at least 16 weeks post-fertilization[21]. Together, these observations support the hypothesis that thereexist probably some meiosis-inhibiting substances (MIS) to sup-press the function of RA or delay germ cells differentiation and pro-mote their mitosis in human and chicken but not in mice. A previousstudy showed that while retinoids induce differentiation in manycellular contexts, f ibroblast growth factors (FGFs) and the FGF sig-naling can maintain an undifferentiated cell state [31].

The basic FGF (bFGF or FGF2) is a member of the FGF family thatplays diverse roles in regulating cell proliferation, migration anddifferentiation during embryonic development [3,4,43]. As animportant cytokine, bFGF is thought to play an essential role inthe reproductive system. Several reports indicated that bFGF is amitogen for primordial germ cells (PGC) and plays a pivotal role

174 B. He et al. / General and Comparative Endocrinology 176 (2012) 173–181

in ovarian function, especially in follicular development [11,12,25,35,36,39]. Earlier experiments demonstrate that bFGF is a stimulusfor proliferation of chicken PGCs, follicular granulosa and thecacells in vitro [11,25,36]. Although bFGF has been shown to stimu-late proliferation of the cultured PGCs, there is no evidence forits role in the early post-implantation embryos at the time germcell allocation occurs [39]. Resnick et al. [35] reported that ovariangerm cells bound to bFGF but reduced to nonspecific levels by day13.5 post coitum (dpc) when most oogonia were entering meioticprophase I in mice. In the immature testes, bFGF is mostly ex-pressed by the sertoli cells, and then its mRNA level is downregu-lated in adult testis [9,15]. It is interesting that bFGF transcriptionis downregulated when germ cells enter meiosis. Previous studiesindicate that bFGF can maintain various cells at undifferentiatedstate. For example, in chickens, bFGF restricts the pool of photore-ceptor cells in favor of cells of the inner retina, increases andmaintains their precursor pool, delays their differentiation [13].Nevertheless, the regulation of ovarian germ cell by bFGF is notclear.

In the present study, we hypothesized that bFGF may suppressmeiosis and promote mitosis of ovarian germ cells, hence enlargethe number of germ cells. Therefore, the expression of bFGF inembryonic chicken ovary during meiosis progression was ana-lyzed. Ovaries from E12.5 embryos and germ cells from E18 werecultured to determine whether bFGF may act directly on germ cellsin the process of mitosis to meiosis switching. In addition, parallelin vivo experiment was carried out to investigate the direct effectof bFGF on germ cell development by injection into the embryos.The results will help to delineate the role of bFGF in regulating longoogonia proliferating phase and the occurrence of an asynchronousmeiosis initiation in the chicken ovary.

Table 1Primers for PCR analysis.

Gene Accession No. Primer sequence (50 to 30) Productlength (bp)

Stra8 XM_416179 GTGAGGGACAGTGGAGGTAACAGAAATGCCGCTTGTAA AT

166

Scp3 XM_416330 CTGTATTTCAGCAGTGGGATGTGCGAAGTTCATTTTGTGC

225

bFGF NM_205433 GGCACTGAAATGTGCAACAGTCCAGGTCCAGTTTTTGGTC

151

18S rRNA AF_173612 CACGGACAGGATTGACAGATGGACATCTAAGGGCATCACA

234

FGFR1 NM_205510 GGAGCGAGACCACCTACTTCGGCATAGCGGACCTTGTAC

308

FGFR2 NM_205319 GAACTCCAACACGCCTCGGGACCCTGTTAATATCA

500

FGFR3 NM_205509 ACCCAACTCCCACCATTTACTTGTCCCATCATCAGGTCCATACTT

368

FGFR4 AF_083063 AGCCCGTCTACGTGCACAGTAGTTGCCGCGGTCGGA

251

Dazl NM_204218 CTGGGGAGCAAAGAAACTACGCAAAGGTGTTCCTCAGACGGT

213

Dmc1 XM_425477 AGATGACAACAAGACGAGCACTCAATCCCACCACCCAGAA

195

Raldh2 NM_204995 GGCAGTTCTTGCTACTATGGATCTGCCCAACCAGCGTAAT

144

Cyp26b1 XM_426366 TTCGCCTCTTCACCCCCATTGGAACCTGCCCTCCTTGTC

225

RARa NM_204536 TGCTTCGAGGTCGGGATGCTGTTGTTCGTGGTGTATTTGC

197

RARb NM_205326 CCTCGTGTTCACCTTTGCAGGCTTGTTGGGTCGTCT

193

RARc X73973 TCATCAAGATCGTGGAGTTCGTGCGGTTCAGCGTCAGC

177

2. Materials and methods

2.1. Animals

Fertilized Hyline chicken (Gallus gallus) eggs were obtainedfrom a commercial hatchery and incubated in an egg incubatorat 38.5 �C and 60% humidity. All procedures were performed inaccordance with the Guiding Principles for the Care and Use of Lab-oratory Animals of Zhejiang University.

2.2. Organ culture

Individual left ovaries of the E12.5 embryos without the meso-nephros were removed for organ culture. Ovaries were cultured onMillipore filters (pore size: 0.45 lm) as previously described for ratovaries [27]. Each ovary was placed into a well of a 24-well platecontaining 500 ll DMEM (Hyclone, Utah, USA) supplemented with10 lg/ml insulin, 5 lg/ml transferrin and 30 nM selenite (Sigma–Aldrich, St. Louis, MO, USA) as the complete medium (ITS medium).The medium also included 2 mM glutamine, 1.75 mM HEPES,100 IU/ml penicillin and 100 lg/ml streptomycin (Sangon, Shang-hai, China). After 48, 72 or 96 h culture at 38.5 �C/5% CO2, the ova-ries were fixed in 4% paraformaldehyde in PBS (30 min at roomtemperature) for fluorescence immunohistochemistry or Giemsastaining. At the beginning of culture, the cultured ovaries werechallenged with 25 ng/ml bFGF (Stemcell Technologies Inc.,Burnaby, BC, Canada) or 2 lM all-trans RA (Sigma–Aldrich).

2.3. Culture of ovarian cells

The procedures of dispersion and culture of ovarian cells werecarried out according to a previous method [41]. Briefly, left ova-ries from E18 embryos were minced and digested twice with

1 mg/ml collagenase (GIBCO BRL, CA, USA) at 37 �C in a shakingwater bath for 15 and 10 min, respectively. The dispersed cellswere filtrated through a 150-lm mesh and then seeded in colla-gen-treated 96-well culture plates at a density of 5 � 104/well in200 ll McCoy’s 5A medium (GIBCO BRL) with ITS. Cells were incu-bated at 38.5 �C under a water-saturated atmosphere containing95% air and 5% CO2 for 48 h. The cultured cells were treated withbFGF (0.1–100 ng/ml), a FGFR1 antagonist SU5402 (1–100 lM,Calbiocherm, La Jolla, CA, USA), forskolin (FRSK, 1 lM, a PKA activa-tor), H89 (0.01–1 lM, a PKA inhibitor), phorbol 12-myristate 13-acetate (PMA, 10 nM, a PKC activator) and H7 (0.01–1 lM, a PKCinhibitor, Sigma–Aldrich) alone or in combinations. In all cases,the control received the vehicle only. Cells were identified andcounted according to their size, as the diameter of germ cellswas between 15 and 25 lm, greater than somatic cells.

2.4. Injection of bFGF into embryos

Injections were made directly into the albumen at the pointedend of the fertilized Buff chicken eggs with 5 lg bFGF (Biopharma-ceutical R&D Center of Jinan University, Guangzhou, China) in 50 llper embryo on E15, E17 and E19 of incubation, respectively. Thecontrol eggs were injected with 50 ll saline only. Left ovaries werecollected from newly-hatched chicks and fixed in 4% neutral para-formaldehyde overnight for paraffin section. The sections of 5 lmthickness were stained with hematoxylin and eosin. The morpho-logical changes were observed under Eclipse 80i microscope (Ni-kon, Japan) and the pictures were captured with a digital camera(DS-Fi1, Nikon, Japan).

2.5. RNA extraction and RT-PCR

Total RNA was extracted from ovaries of E12.5 to E16.5 embryosor cultured ovaries for 24 or 48 h with Trizol reagent (Invitrogen

B. He et al. / General and Comparative Endocrinology 176 (2012) 173–181 175

Co., Carlsbad, CA, USA). The RNA purity and concentration weredetermined spectrophotometrically at 260/280 nm in the rangeof 1.8–2.0. Total RNA (2 lg) was reverse transcribed by FermentasOne step RT-PCR kit (MBI Fermentas, Burlington, ON, Canada)according to the manufacture’s instruction and further amplifiedby PCR. The sequences of the primers were listed in Table 1. PCRamplification was performed on a 25 ll volume containing 1 llRT products. The PCR products were analyzed by electrophoresison 1.5% agarose gel and quantified with Tanon Gel Imaging system(Tanon, Shanghai, China).

2.6. Real time PCR analysis

Real time RT-PCR (qRT-PCR) was used to assess the expressionof bFGF, pre-meiotic marker Stra8 (stimulated by retinoic acidgene), meiotic markers Scp3 (a gene encoding a synaptonemalcomplex protein) and Dmc1 (a gene encoding a meiotic recombi-nase), RA metabolism-related enzymes Raldh2 (a gene encodingretinaldehyde dehydrogenase, type 2, also called as Aldh1a2) andCyp26b1 (a gene encoding a RA metabolizing cytochrome P450 en-zyme) and RA receptors RARa, RARb and RARc (Table 1). The qPCRwas carried out on ABI 7300 HT real time PCR machine (AppliedBiosystems, Foster City, CA, USA) with the reaction volume of20 ll consisting of cDNA from 10 ng of original RNA template,400 nM of each of the gene-specific forward and reverse primersand 10 ll SYBR� Premix Ex Taq™ (TaKaRa Bio Inc., Japan). Individ-ual samples were analyzed in triplicate and experiments were per-formed twice. All samples were normalized against Dazl (a geneencoding a germ-cell-specific RNA binding protein) or 18S rRNAusing the comparative CT method (DDCT).

2.7. In situ hybridization

Specific primers used for the production of Digoxigenin-labeledprobes for detecting FGFR1 mRNA were as follows: 5-GAAGTGCCTCCATCTTCTGGGCTGGTGCTGGT-3. For section in situ hybridiza-tion, tissues were embedded in paraffin, sectioned at 5 lm. Thesections were deparaffinized and permeabilised with a high con-centration of proteinase K in (3% sodium citrate) for 15–20 minat 37 �C. Then acetylated and incubated overnight at 38–42 �C withhybridization buffer containing a Digoxigenin-labeled probes for12 h. Hybridized sections were washed at 37 �C in succession with2 � SSC (saline sodium citrate), 0.5 � SSC, and 0.2 � SSC, blockedwith 5% BSA in buffered saline (PBS) for 30 min, and incubatedwith an biotin-labeled anti-Digoxigenin. After brief washing withPBS, Then it was detected by using SABC system, as described inthe manufacturer’s protocol (Boster Bioengineering Co., Ltd.,Wuhan, China). Tissues or sections were mounted in 70% glyceroland observed using an Eclipse 80i microscope (Nikon, Japan) andthe pictures were captured with a digital camera (DS-Fi1, Nikon,Japan).

2.8. Fluorescence immunohistochemistry

Fluorescence immunohistochemistry was performed with fixedovaries immersed in OCT embedding compound and snap frozen.Ten micron cryosections were permeabilised in 1% Triton X-100/TBS for 10 min and then incubated with primary antibodies over-night at 4 �C. Sources and dilution of primary antibodies were asfollows: rabbit anti-bFGF antibody (1:100, SC-79, Santa Cruz Bio-technology, Inc., CA, USA), mouse anti-Scp3 antibody (1:500,ab97672, Abcam, Cambridge, UK), mouse anti-SSEA1 antibody(1:500, MC-480, Developmental Studies Hybridoma Bank (DSHB),Iowa, USA) or rabbit anti-c-Kit antibody (1:200, BA0467, BosterBioengineering Co., Ltd., Wuhan, China). TRITC-labeled goat anti-mouse IgM (1:1000, Santa Cruz, UC, USA), FITC-labeled goat anti-

rabbit IgG or FITC-labeled goat anti-mouse IgG (1:1000, KPL Inc.,Gaithersburg, MD, USA) was used as the second anti-body. Sectionswere counterstained with DAPI (300 nM, Sigma–Aldrich). Negativecontrol sections were incubated with normal serum instead of pri-mary antibodies.

2.9. BrdU incorporation assay

After 36 h treatment with bFGF, 25 lg/ml BrdU (Sigma–Aldrich)was added into ovarian culture medium and the incubation wascontinued for additional 12 h for incorporation. Frozen ovarian sec-tions were performed with BrdU fluorescence immunohistochem-istry according to a previous method [14] with mouse anti-BrdUmonoclonal antibody (1:500, G3B4; DSHB) and subsequently withFITC-labeled goat anti-mouse IgG as the second antibody.

2.10. Giemsa staining

The cultured ovaries and left ovaries of newly-hatched chickswas incubated in 1.5 ml of 0.25% trypsin (GIBCO BRL) at 37 �C ina shaking water bath for 30 min and was completely dissociatedwith a siliconized Pasteur pipette. The sedimented cells wereresuspended in 0.56% KCl solution (hypotonic solution) and incu-bated at 37 �C for 30 min. The suspension was centrifuged againthe cells were resuspended in a chilled fixative at 4 �C (3:1 mixtureof absolute methyl alcohol and glacial acetic acid). After 1 h, thischilled suspension was centrifuged and resuspended again in afreshly prepared chilled fixative. Finally, the slides were preparedaccording to the air-drying method. Cell smears were stained withGiemsa. The germ cells of each ovary were classified as follows:oogonia at interphase; metaphase of mitosis; oocytes at the prelep-totene, leptotene, zygotene, pachytene, and diplotene phases ofmeiosis. Germ cells were recognized by their relatively larger size.The percentage of variations in oogonia or oocytes at each stage ofmeiotic prophase were evaluated directly by differential counts inthe squash preparations.

2.11. Statistical analysis

All data were expressed as the means ± SEM and analyzed byANOVA and Duncan’s multiple-range tests using the SAS 8.0 soft-ware. P < 0.05 was considered to be statistically different.

3. Results

3.1. bFGF and FGFR mRNAs expression during ovarian development

The pre-meiotic and meiotic germ cell marker Stra8 and Scp3were determined during ovarian development at the crucial periodof female germ cell meiosis between E12.5 and E16.5. The resultsshowed that Stra8 mRNA expression started to increase at E12.5and Scp3 mRNA expression started to increase at E14.5 (Fig. 1A).On the contrary, bFGF mRNA expression remained at higher levelfrom E12.5 to E13.5 in ovaries, subsequently its abundance beganto drop at E14.5 (Fig. 1A). No significant change of bFGF mRNAexpression in testis (Fig. 1A). The expression of bFGF mRNA mani-fested an inverse relationship with the meiotic marker genes. Theexpression of bFGF protein was also revealed by immunohisto-chemistry with specific antibodies. The bFGF showed cytoplasmicstaining in both ovarian somatic and germ cells in E12.5 (Fig. 1B-a) and E16.5 (Fig. 1B-b). The bFGF-positive cells decreased in theovarian cortex at E16.5 (Fig. 1B-b). In addition, expression of threeFGF receptors: FGFR1, FGFR2 and FGFR3 were revealed in E12.5 ova-ries (Fig. 2A). In situ hybridization analysis manifested FGFR1

Fig. 1. Expression of bFGF at different stages of chicken gonad development, assessed by qRT-PCR and fluorescence immunohistochemistry. (A) Analysis of Stra8, Scp3 andbFGF mRNAs expression from E12.5 to E16.5 by qRT-PCR. Dazl mRNA was used as the normalization control for Stra8 and Scp3, and 18S rRNA was used for bFGF mRNAexpression analysis. Values are means ± SEM of three experiments. (B) Detection of bFGF protein (green) in cytoplasm of E12.5 (a) and E16.5 (b) ovarian somatic and germcells. Scale bar: 20 lm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this paper.)

176 B. He et al. / General and Comparative Endocrinology 176 (2012) 173–181

mRNA expression in the cytoplasm of E12.5 (Fig. 2B-a) and E18(Fig. 2B-b) ovarian germ cells.

3.2. Effect of RA on meiosis initiation

As an efficient inducer of meiosis, RA treatment at 2 lM signifi-cantly elevated expression of Stra8 (133% increase), Scp3 (44% in-crease) and Dmc1 (227% increase) mRNAs in cultured E12.5ovaries (Fig. 3A–C). This stimulation was further confirmed by di-rect observation of meiosis marker Scp3 fluorescence immunohis-tochemistry on cryosections (Fig. 4). Treatment with 2 lM RA for96 h increased the percentage of oocytes using Giemsa staining(Fig. 5) in the cultured ovaries. The percentage of oocytes at lepto-tene and zygotene was increased 1.4-fold and 2.4-fold, respectively.

3.3. Effect of bFGF on RA-induced meiosis

After treatment of the cultured E12.5 ovaries with 25 ng/mlbFGF alone for 48 h, both Scp3 and Dmc1 mRNA expression was sig-nificantly decreased compared with the control (Fig. 3B and C).Though RA treatment induced higher expression of Scp3 andDmc1 mRNAs, simultaneous treatment with bFGF remarkably de-creased this stimulation in the cultured ovaries (Fig. 3B and C). Un-like Scp3 and Dmc1 mRNAs, the expression of Stra8 mRNAdisplayed a different pattern in which bFGF did not elicit signifi-cant changes in the basal or RA-stimulated expressed after 24 htreatment (Fig. 3A). By fluorescence immunohistochemistry, treat-ment with bFGF decreased Scp3-positive cells that were elevated

by RA (Fig. 4). In addition, treatment with bFGF for 96 h decreasedthe percentage of oocytes (8.1% decrease at leptotene, 2.3% de-crease at zygotene and 1.9% decrease at pachytene, respectively)through Giemsa staining (Fig. 5). Meanwhile, expression of severalenzymes crucial for RA metabolism was examined. Treatment ofthe cultured ovaries by bFGF did not cause significant changes inthe mRNA of the RA synthesizing enzyme Raldh2 transcriptionand the RA degrading enzyme Cyp26b1 transcription over a 24 hperiod (Fig. 3D). Furthermore, no significant change of RA receptors(RARa, RARb and RARc) mRNAs was displayed by qRT-PCR aftertreatment with bFGF (Fig. 3D).

3.4. bFGF promotes germ cell mitosis

Parallel with the meiosis-inhibiting activity, treatment of thecultured E12.5 ovaries with 25 ng/ml bFGF increased germ cellproliferation in the ovarian cortex after 48 h. The BrdU/SSEA1 dou-ble positive cells were remarkably increased in the bFGF-treatedovaries (Fig. 6). Besides the in vitro action of bFGF on cultured ova-ries, the proliferating effect of bFGF was further confirmed byin vivo injection study. The thickness of the newly-hatched chickovarian cortex was significantly elevated by 34.7% after bFGF treat-ment (Fig. 7A) while the hatchability remained almost unchanged(P > 0.05). More importantly, injection of bFGF caused significantaccelerated germ cell proliferation. This increase in germ cell num-ber resulted dominantly from the increase in oogonia number(Fig. 7B) instead of the oocytes. The proportion of undifferentiated

Fig. 2. Expression of FGFRs mRNAs in E12.5 and E18 in embryonic chicken ovaries,assessed by RT-PCR and in situ hybridization. (A) Expression of FGFR1, FGFR2 andFGFR3 mRNAs in E12.5 ovaries detected by RT-PCR. (B) FGFR1 transcripts aredetected in cytoplasm of E12.5 (a) and E18 (b) ovarian germ cells and somatic cells.Scale bar: 20 lm.

Fig. 4. Effect of RA and bFGF on meiosis in cultured E12.5 ovaries, assessed by Scp3fluorescence immunohistochemistry. Ovaries were cultured for 72 h with 2 lM RA,25 ng/ml bFGF or both, respectively. Scale bars: 50 lm.

B. He et al. / General and Comparative Endocrinology 176 (2012) 173–181 177

germ cells (oogonia) in bFGF-injected individuals (44.1%) was sig-nificantly higher than the control (26.0%).

Beside the ovarian organ culture, dispersed ovarian cell culturewas used to reveal the proliferating effect of bFGF on germ cells.Germ cells (oogonia) E18 chicken ovaries were positive for c-Kitand the somatic cells were negative (Fig. 8A). The germ cells pos-sessed clear three-dimensional oval appearance after 10 ng/mlbFGF treatment, while the cells of the control displayed flat form.Moreover, bFGF-stimulated cell proliferation manifested a dose-dependent manner from 0.1 to 100 ng/ml after 48 h treatment(Fig. 8B).

Fig. 3. Effect of RA and bFGF on meiosis in cultured E12.5 ovaries. Ovaries were cultured wand expression of Scp3 (B) and Dmc1 (C) mRNAs cultured for 48 h were analyzed. (D) Expovaries with 25 ng/ml bFGF for 24 h. Dazl was used as a reference in A–C and 18S rRNA instatistically different. The ns means no significant difference between different treatme

3.5. Involvement of FGFR1 signaling and PKC activation in bFGF-stimulated cell proliferation

Treatment of 10 ng/ml bFGF in cultured germ cells for 48 h wasadopted for subsequent studies. The pro-proliferation action ofbFGF on germ cells was significantly inhibited by SU5402 at 1–100 lM (Fig. 8C). As a PKC activator, PMA treatment promoted pro-liferation of germ cells. On the contrary, inhibition of PKC with H7

at 0.01–1 lM attenuated the bFGF-stimulated cell proliferation(Fig. 8D). Compared with the PKC system, activation of PKA byFRSK or inhibition by H89 did not change the bFGF-stimulated cellproliferation significantly (Fig. 8E).

4. Discussion

Successful gametogenesis require accurate interplay of hor-mones, growth factors, cytokines, microRNA and other substances

ith 2 lM RA, 25 ng/ml bFGF, or both. Expression of Stra8 mRNA cultured for 24 h (A)ression of Cyp26b1, Raldh2, RARa, RARb and RARc mRNAs were analyzed in culturedE. Values are means ± SEM of three experiments. Bars with different superscripts arent (P < 0.05).

Fig. 5. Squash preparations of the left ovaries of newly-hatched chicks. (A) Oogoniaat interphase (a) and metaphase (b) of mitosis, oocytes at preleptotene (c), oocyte atleptotene (d), oocyte at zygotene (e), oocytes at pachytene (f), oocytes at diplotenewith somatic cells (g), and oogonia undergoing cell division (h). Scale bar: 5 lm. (B)Stage of meiosis prophase I as determined by Giemsa stain after 96 h of culture with2 lM RA and 25 ng/ml bFGF or both, respectively. Values are means ± SEM of threeexperiments. Asterisk means statistically different between different treatments(P < 0.05).

Fig. 6. BrdU incorporation of germ cells in ovaries after 25 ng/ml bFGF treatmentfor 48 h. Arrow: BrdU-positive cells; Arrowhead: SSEA-1 positive cells. Scale bar:20 lm.

178 B. He et al. / General and Comparative Endocrinology 176 (2012) 173–181

that orchestrate germ cells to finish mitosis and meiosis, then formhaploid gametes [5,6,16,17,18,24]. Studies dealing with avian spe-cies have demonstrated that the meiotic initiation of oogonia occurin the embryonic ovary of birds and cease at hatching [37]. Here,we adopted ovarian organ and cell culture methods, together within vivo method to examine the actions of a cytokine bFGF on chick-en germ cells mitosis and meiosis progression. The results showthat bFGF plays dual roles in female germ cells development inthe embryonic chicken: as a MIS to suppress meiosis and as a mito-gen to promote mitosis.

In the ovarian organ culture from E12.5 embryos, pre-meioticand meiotic marker genes including Stra8, Scp3 and Dmc1 mRNAswere upregulated specifically after RA treatment for 24–96 h as as-sessed by different methods including qRT-PCR, immunohisto-chemistry and Giemsa staining. This is in accordance withprevious reports that in chickens, pre-meiotic DNA synthesis beganin the germ cells of the left ovary between E15 and E16 [8,37].Since meiosis is unlikely autonomous for the germ cells committedto the female pathway, it is widely accepted that meiosis is stimu-lated by RA which is synthesized in the somatic cells [1,5,23].

While retinoids can drive differentiation in many different con-texts, FGF signaling can maintain an undifferentiated cell state[31]. In our result, bFGF were detected in cytoplasm of both ovar-ian somatic and germ cells from E12.5 to E16.5 embryos. OvarianbFGF mRNA expression was downregulated in E14.5 when Scp3mRNA expression increasing occurs. The bFGF-positive cells weredecreased in cortex of ovary. It indicates that bFGF was expressedin germ cells and then downregulation of bFGF occurs as a result ofenter meiosis. Expression of bFGF in germ cells has been describedby previous reports and played a role in spermatogenesis [9,15].Alternatively, bFGF produced by somatic cells bound to germ cellsin the ovary to inhibit meiosis in pre-meiosis phase in chickensthrough auto/paracrine action via their respective receptors. As aheparin-binding growth factor, bFGF is known to exert diversephysiological functions by binding and activating FGFR that is en-coded by four distinct genes, FGFR1–4 [33]. In this study, three FGFreceptors: FGFR1, FGFR2 and FGFR3 were expressed in E12.5 ova-ries. All three receptors can bind with bFGF which makes the actionof bFGF plausible in the ovary. Also, FGFR1 transcripts are detectedin cytoplasm of E12.5 ovarian cortex germ cells. In the mouse, bFGFbinding to ovarian germ cells was reduced to nonspecific level by13.5 dpc when most oogonia are entering meiotic [35]. To clarifythis hypothesis, we added bFGF into cultured ovaries, baseline lev-els of meiotic markers gene expression were significantly lowerthan the control. Also, the number of Scp3-positive cells and pro-portion of meiotic germ cells were decreased after treatment bybFGF. A corollary of these findings is that while RA acts to pushgerm cells toward an oogenesis fate, bFGF acts to retard their entryinto meiosis.

In this study, RA triggers germ cells to enter meiosis and bFGFsuppresses meiotic genes Scp3 and Dmc1 mRNA expression butnot Stra8, which is activated by RA. In the next experiment, weexamined if bFGF works on expression of components for RA syn-thesis, metabolism and signaling pathways. Available data indicatethat Raldh2 is the major enzyme to catalyze RA synthesis andCyp26b1 is the major enzyme to degrade RA in the avian embryo[34,38]. In addition, RA exerts its action by binding to the nuclearreceptors RAR and RXR to regulate gene expression profile in thetarget cells. With qRT-PCR, we found no significant difference inmRNA abundance of Raldh2, Cyp26b1 and genes encoding the RAreceptors (RARa, RARb and RARc) between bFGF-treated and con-trol ovaries. These results, taken together with the expression pat-terns between bFGF and Stra8 in embryonic period indicate thatbFGF controls meiosis progression likely independent of RA.

Fig. 7. Effect of bFGF on thickness of ovarian cortex, hatchability of chicken and germ cell number after in vivo injection. (A) Thickness of ovarian cortex and hatchability ofchickens. (B) Effect of bFGF on the component of germ cells from chicken ovaries. Values are means ± SEM of three experiments. Bars with different superscripts arestatistically different (P < 0.05).

Fig. 8. Action of FGFR1, PKC and PKA systems in bFGF-induced proliferation of germ cells. (A) The E18 ovarian germ cells were identified by c-Kit immunocytochemicalstaining. Scale bar: 50 lm. (B) Effect of bFGF (0.1–100 ng/ml) on proliferation of germ cells after bFGF treatment for 48 h. (C) Effect of FGFR1 blockade by SU5402 (0–100 lM)on bFGF (10 ng/ml) suppressed proliferation of germ cells after 48 h culture. (D) and (E) Changes in cells number by PKC/PKA activation (PMA, 10 nM/FRSK, 1 lM), andinhibition (H7, 0.01–1 lM/H89, 0.01–1 lM), after treatment for 48 h, respectively. Values are means ± SEM of three experiments. Bars with different superscripts arestatistically different (P < 0.05).

B. He et al. / General and Comparative Endocrinology 176 (2012) 173–181 179

Our findings obviously supports the observations of the earlierstudies that antagonism between FGFs and RA in germ cell devel-opment [4]. But it is not clear why the bFGF antagonise meiosis butnot Stra8 mRNA expression. Another study demonstrated the in-creased Stra8 mRNA expression in response to RA in cultures of hu-man fetal testes [10], but any changes in the levels of Scp3 andDmc1 mRNA were not detected. Combining these data with ourfinding suggest that whilst RA is sufficient to induce Stra8 expres-

sion, it cannot trigger widespread meiosis-associated gene expres-sion in the germ cells. Probably some MIS act to suppress or delaygerm cells meiosis progression. Our result confirms that bFGF ex-ists as a MIS involved in ovarian germ cell meiosis progression in-stead of meiotic initiation.

In a recent study, another member of the FGF family, FGF9makes germ cells less responsive to RA, and forces germ cells tomaintain expression of pluripotency genes and to upregulate male

180 B. He et al. / General and Comparative Endocrinology 176 (2012) 173–181

germ cell fate genes [4]. RA and FGF9 produced locally by the so-matic cells act in opposite ways to regulate the RNA binding pro-tein Nanos2 and meiotic entry during fetal development in themouse [2]. As MIS, both bFGF and FGF9 force germ cells to main-tain at undifferentiated status. FGF9 is initially expressed in gonadsof both sexes, but it becomes male restricted at meiotic initiation[22]. The sexual dimorphism expression of FGF9 plays a key rolein regulating germ cell sexual fate. Our results displayed that bFGFwas expressed in gonads of both sexes, but not markedly higher intestes than in ovaries. By fluorescence immunohistochemistry,bFGF-positive cells were decreased in the ovarian cortex duringmeiosis progression. Therefore, collecting data from our experi-ment could reach a conclusion: bFGF that is produced in germ cellsor binds to germ cells may inhibit or delay germ cell meiosis. Fur-thermore, based on their function on meiosis, our results suggestthat MIS likely consists of two categories, one is inter-sex MIS withsexual dimorphism expression determines germ cell sexual fate,e.g., FGF9 [4] and Nanos2 [2], another is intra-sex MIS without sex-ual dimorphism expression regulates meiosis progression, e.g.,bFGF.

In addition to inhibiting differentiation, a recent study indicatesthat bFGF is one of the key factors that enable proliferation ofchicken PGCs via MEK/ERK signaling [11]. In the present study,we proved that bFGF treatment (25 ng/ml) on cultured ovaries elic-ited a rapid increase in germ cell proliferation by BrdU incorpora-tion. In this study, there was a downregulation of bFGF mRNA inE14.5 ovaries. Therefore, we injected bFGF on E15, E17 and E19of incubation for a complement to examine the action of bFGF onthe undifferentiated germ cells during meiosis phase. Germ cellproliferation was significantly enhanced after bFGF treatment. Aprevious study indicates that there is a subset of oogonia whichstill fail to enter meiosis even at the time of hatching [19], thuswe predict that a cohort of ‘residual’ oogonia may be the targetof bFGF’s proliferative action. Treatment with bFGF may enablethese residual germ cells to accelerate proliferation, thus yieldmore germ cells. A previous report showed that the germ cell pop-ulation as a whole was still proliferating up to E17 of incubation[19]. In this study, bFGF-positive cells were decreased in ovariancortex from E12.5 to E16.5. These results confirm that bFGF is animportant factor to enable proliferation of ovarian germ cellsin vivo.

To elaborate the functions of bFGF on germ cell development,cultured germ cells in monolayer were treated with exogenousbFGF. The result showed that bFGF-stimulated germ cell prolifera-tion was significantly inhibited by an FGFR1 antagonist SU5402.Moreover, FGFR1 transcripts are detected in cytoplasm of E12.5and E18.5 germ cells in the ovarian cortex, which indicates thatbFGF exerts its proliferating action on germ cells by binding withthe cell surface receptor FGFR1. This result clearly suggests thatbFGF may regulate proliferation of germ cells predominantlythrough auto/paracrine action via their respective receptors. Thesignal transduction of bFGF is related with development andgrowth of diverse cell types, which involves PKA and PKC signalingpathways [20,28,32]. Many evidences prove that PKA and PKC iso-zymes play crucial roles in the control of germ cell proliferation[7,26,42]. Activation of PKC by PMA significantly increased thegerm cell number, whereas inhibition of PKC by H7 remarkablyattenuated this effect. However, activation or inhibition of PKAhad no significant effect on the bFGF-stimulated cell proliferation.Taken together, these results suggest that PKC rather than PKA sig-naling pathway is involved in bFGF-stimulated proliferation ofchicken germ cells in embryonic ovaries.

It is generally accepted that the process of germ cell mitosis iscompleted and followed by meiosis in the embryonic mammalianand avian species [44]. This is true for the majority of oocytes.However, there are a small percentage of germ cells have not yet

detected or received the signals to enter meiosis whereas other oo-cytes have already entered meiotic arrest in chicks [19,30]. Here,we found supplementation with bFGF increased the number ofSSEA1-positive cells. This result strongly suggests that bFGF maymaintain germ cells at the undifferentiated status. We consideredthat germ cells which expressed FGFR to bind to bFGF may shieldfrom the action of RA and maintained an undifferentiated statefor self renewal.

In conclusion, this study showed that bFGF in embryonic chick-en ovaries displayed a decreased expression during meiosis pro-gression. With ovarian organ or cell culture in vitro and also byinjection into embryos in vivo, we revealed that bFGF suppressesRA-induced entry of germ cells into meiosis to ensure the germcells to maintain at undifferentiated status and accelerate cell pro-liferation by binding with FGFR1 involving PKC activation. In addi-tion, bFGF is likely involved in maintenance of the oogoniaproliferating phase and the occurrence of an asynchronous meiosisinitiation in embryonic chicken ovary.

Acknowledgments

This study was supported by the National Natural ScienceFoundation of China (No. 30871843), Zhejiang Provincial NaturalScience Foundation (Z3110115) and Chinese Universities ScientificFund. We are grateful to Imdad Leghari (Zhejiang University) forEnglish improvement in the manuscript.

References

[1] A.E. Baltus, D.B. Menke, Y.C. Hu, M.L. Goodheart, A.E. Carpenter, D.G. de Rooij,D.C. Page, In germ cells of mouse embryonic ovaries, the decision to entermeiosis precedes premeiotic DNA replication, Nat. Genet. 38 (2006) 1430–1434.

[2] F. Barrios, D. Filipponi, M. Pellegrini, M.P. Paronetto, S. Di Siena, R. Geremia, P.Rossi, M. De Felici, E.A. Jannini, S. Dolci, Opposing effects of retinoic acid andFGF9 on Nanos2 expression and meiotic entry of mouse germ cells, J. Cell Sci.123 (2010) 871–880.

[3] A. Beenken, M. Mohammadi, The FGF family: biology, pathophysiology andtherapy, Nat. Rev. Drug Discov. 8 (2009) 235–253.

[4] A. Bikfalvi, S. Klein, G. Pintucci, D.B. Rifkin, Biological roles of fibroblast growthfactor-2, Endocr. Rev. 18 (1997) 26–45.

[5] J. Bowles, D. Knight, C. Smith, D. Wilhelm, J. Richman, S. Mamiya, K. Yashiro, K.Chawengsaksophak, M.J. Wilson, J. Rossant, H. Hamada, P. Koopman, Retinoidsignaling determines germ cell fate in mice, Science 312 (2006) 596–600.

[6] J. Bowles, C.W. Feng, C. Spiller, T.L. Davidson, A. Jackson, P. Koopman, FGF9suppresses meiosis and promotes male germ cell fate in mice, Dev. Cell 19(2010) 440–449.

[7] M.J. Bragado, I.M. Aparicio, M.C. Gil, L.J. Garcia-Marin, Protein kinases A and Cand phosphatidylinositol 3 kinase regulate glycogen synthase kinase-3A serine21 phosphorylation in boar spermatozoa, J. Cell. Biochem. 109 (2010) 65–73.

[8] M. Callebaut, Premeiosis and premeiotic DNA synthesis in the left ovary of thefemale chick embryo, J. Embryol. Exp. Morphol. 18 (1967) 299–304.

[9] B. Cancilla, A. Davies, M. Ford-Perriss, G.P. Risbridger, Discrete cell- and stage-specific localisation of fibroblast growth factors and receptor expressionduring testis development, J. Endocrinol. 164 (2000) 149–159.

[10] A.J. Childs, G. Cowan, H.L. Kinnell, R.A. Anderson, P.T. Saunders, Retinoic acidsignalling and the control of meiotic entry in the human fetal gonad, PLoS One6 (2011) e20249.

[11] J.W. Choi, S. Kim, T.M. Kim, Y.M. Kim, H.W. Seo, T.S. Park, J.W. Jeong, G. Song,J.Y. Han, Basic fibroblast growth factor activates MEK/ERK cell signalingpathway and stimulates the proliferation of chicken primordial germ cells,PLoS One 5 (2010) e12968.

[12] G. Durcova-Hills, I.R. Adams, S.C. Barton, M.A. Surani, A. McLaren, The role ofexogenous fibroblast growth factor-2 on the reprogramming of primordialgerm cells into pluripotent stem cells, Stem Cells 24 (2006) 1441–1449.

[13] F. Frohns, M. Mager, P.G. Layer, Basic fibroblast growth factor increases theprecursor pool of photoreceptors, but inhibits their differentiation andapoptosis in chicken retinal reaggregates, Eur. J. Neurosci. 29 (2009) 1931–1942.

[14] C. Ge, M. Yu, J.N. Petitte, C. Zhang, Epidermal growth factor-inducedproliferation of chicken primordial germ cells: involvement of calcium/protein kinase C and NFKB1, Biol. Reprod. 80 (2009) 528–536.

[15] I.G. Gonzalez-Herrera, L. Prado-Lourenco, F. Pileur, C. Conte, A. Morin, F. Cabon,H. Prats, S. Vagner, F. Bayard, S. Audigier, A.C. Prats, Testosterone regulatesFGF-2 expression during testis maturation by an IRES-dependent translationalmechanism, FASEB J. 20 (2006) 476–478.

B. He et al. / General and Comparative Endocrinology 176 (2012) 173–181 181

[16] M.G. Gonzalez-Moran, Effect of follicle-stimulating hormone on different cellsub-populations in the ovary of newly hatched chicks treated duringembryonic development, Br. Poult. Sci. 39 (1998) 128–132.

[17] M.G. Gonzalez-Moran, Effects of luteinizing hormone treatment on oogenesisin ovarian germ cells of the chick (Gallus domesticus), Domest. Anim.Endocrinol. 33 (2007) 154–166.

[18] C.A. Hogarth, D. Mitchell, R. Evanoff, C. Small, M. Griswold, Identification andexpression of potential regulators of the mammalian mitotic-to-meiotictransition, Biol. Reprod. 84 (2011) 34–42.

[19] G.C. Hughes, The population of germ cells in the developing female chick, J.Embryol. Exp. Morphol. 11 (1963) 513–536.

[20] H.J. Im, P. Muddasani, V. Natarajan, T.M. Schmid, J.A. Block, F. Davis, A.J. vanWijnen, R.F. Loeser, Basic fibroblast growth factor stimulates matrixmetalloproteinase-13 via the molecular cross-talk between the mitogen-activated protein kinases and protein kinase Cdelta pathways in human adultarticular chondrocytes, J. Biol. Chem. 282 (2007) 11110–11121.

[21] C.L. Kerr, C.M. Hill, P.D. Blumenthal, J.D. Gearhart, Expression of pluripotentstem cell markers in the human fetal ovary, Hum. Reprod. 23 (2008) 589–599.

[22] Y. Kim, A. Kobayashi, R. Sekido, L. DiNapoli, J. Brennan, M.C. Chaboissier, F.Poulat, R.R. Behringer, R. Lovell-Badge, B. Capel, Fgf9 and Wnt4 act asantagonistic signals to regulate mammalian sex determination, PLoS Biol. 4(2006) e187.

[23] J. Koubova, D.B. Menke, Q. Zhou, B. Capel, M.D. Griswold, D.C. Page, Retinoicacid regulates sex-specific timing of meiotic initiation in mice, Proc. Natl. Acad.Sci. USA 103 (2006) 2474–2479.

[24] S.I. Lee, B.R. Lee, Y.S. Hwang, H.C. Lee, D. Rengaraj, G. Song, T.S. Park, J.Y. Han,MicroRNA-mediated posttranscriptional regulation is required for maintainingundifferentiated properties of blastoderm and primordial germ cells inchickens, Proc. Natl. Acad. Sci. USA 108 (2011) 10426–10431.

[25] J. Lin, Y. Jia, W. Zeng, Y. Mi, C. Zhang, Basic FGF promotes proliferation ofovarian granulosa cells in the laying chickens via FGFR1 and PKC pathway,Reprod. Domest. Anim. 47 (2012) 135–142.

[26] H. Liu, W. Zeng, A. Cao, C. Zhang, Follicle-stimulating hormone promotesproliferation of cultured chicken ovarian germ cells through protein kinases Aand C activation, J. Zhejiang Univ.-Sci. B 11 (2010) 952–957.

[27] G. Livera, V. Rouiller-Fabre, J. Valla, R. Habert, Effects of retinoids on themeiosis in the fetal rat ovary in culture, Mol. Cell. Endocrinol. 165 (2000) 225–231.

[28] T. Maffucci, C. Raimondi, S. Abu-Hayyeh, V. Dominguez, G. Sala, I. Zachary, M.Falasca, A phosphoinositide 3-kinase/phospholipase Cgamma1 pathwayregulates fibroblast growth factor-induced capillary tube formation, PLoSOne 4 (2009) e8285.

[29] A. McLaren, The developmental history of female germ cells in mammals, Oxf.Rev. Reprod. Biol. 10 (1988) 162–179.

[30] M. Naito, T. Minematsu, T. Harumi, T. Kuwana, Testicular and ovariangonocytes from 20-day incubated chicken embryos contribute to germlinelineage after transfer into bloodstream of recipient embryos, Reproduction 134(2007) 577–584.

[31] K. Niederreither, P. Dolle, Retinoic acid in development: towards an integratedview, Nat. Rev. Genet. 9 (2008) 541–553.

[32] C. O’Driscoll, D. Wallace, T.G. Cotter, bFGF promotes photoreceptor cellsurvival in vitro by PKA-mediated inactivation of glycogen synthase kinase3beta and CREB-dependent Bcl-2 up-regulation, J. Neurochem. 103 (2007)860–870.

[33] C.J. Powers, S.W. McLeskey, A. Wellstein, Fibroblast growth factors, theirreceptors and signaling, Endocr. Relat. Cancer 7 (2000) 165–197.

[34] S. Reijntjes, E. Gale, M. Maden, Expression of the retinoic acid catabolisingenzyme CYP26B1 in the chick embryo and its regulation by retinoic acid, GeneExp. Patterns 3 (2003) 621–627.

[35] J.L. Resnick, M. Ortiz, J.R. Keller, P.J. Donovan, Role of fibroblast growth factorsand their receptors in mouse primordial germ cell growth, Biol. Reprod. 59(1998) 1224–1229.

[36] R.D. Roberts, R.C. Ellis, Mitogenic effects of fibroblast growth factors onchicken granulosa and theca cells in vitro, Biol. Reprod. 61 (1999) 1387–1392.

[37] C.A. Smith, K.N. Roeszler, J. Bowles, P. Koopman, A.H. Sinclair, Onset of meiosisin the chicken embryo; evidence of a role for retinoic acid, BMC Dev. Biol. 8(2008) 85.

[38] E.C. Swindell, C. Thaller, S. Sockanathan, M. Petkovich, T.M. Jessell, G. Eichele,Complementary domains of retinoic acid production and degradation in theearly chick embryo, Dev. Biol. 216 (1999) 282–296.

[39] Y. Takeuchi, K. Molyneaux, C. Runyan, K. Schaible, C. Wylie, The roles of FGFsignaling in germ cell migration in the mouse, Development 132 (2005) 5399–5409.

[40] D.J. Wolgemuth, S.S. Roberts, Regulating mitosis and meiosis in the male germline: critical functions for cyclins, Philos. Trans. R. Soc. Lond. B Biol. Sci. 365(2010) 1653–1662.

[41] M. Xie, C. Zhang, Estrogenic and toxic effects of polychlorinated biphenyls oncultured ovarian germ cells of embryonic chickens, Reprod. Toxicol. 19 (2004)79–86.

[42] M. Yu, K. Guan, C. Zhang, The promoting effect of retinoic acid on proliferationof chicken primordial germ cells by increased expression of cadherin andcatenins, Amino Acids 40 (2011) 933–941.

[43] W. Zhou, X. Feng, Y. Wu, J. Benge, Z. Zhang, Z. Chen, FGF-receptor substrate 2functions as a molecular sensor integrating external regulatory signals into theFGF pathway, Cell Res. 19 (2009) 1165–1177.

[44] S. Zuckerman, T.G. Baker, The development of the ovary and the process ofoogenesis, in: S. Zuckerman, B.J. Weir (Eds.), Ovary, Academic Press, New York,1977, pp. 41–67.