The female germline comprises a reserve population of primordial (non-growing) follicles containing diplotene oocytes arrested in the first meiotic prophase. By convention, the reserve is established when all individual oocytes are enclosed by granulosa cells. This commonly occurs prior to or around birth, according to species. Histologically, the ‘reserve’ is the number of primordial follicles in the ovary at any given age and is ultimately depleted by degeneration and progression through folliculogenesis until exhausted. How and when the reserve reaches its peak number of follicles is determined by ovarian morphogenesis and germ cell dynamics involving i) oogonial proliferation and entry into meiosis producing an oversupply of oocytes and ii) large-scale germ cell death resulting in markedly reduced numbers surviving as the primordial follicle reserve. Our understanding of the processes maintaining the reserve comes primarily from genetically engineered mouse models, experimental activation or destruction of oocytes, and quantitative histological analysis. As the source of ovulated oocytes in postnatal life, the primordial follicle reserve requires regulation of i) its survival or maintenance, ii) suppression of development (dormancy), and iii) activation for growth and entry into folliculogenesis. The mechanisms influencing these alternate and complex inter-related phenomena remain to be fully elucidated. Drawing upon direct and indirect evidence, we discuss the controversial concept of postnatal oogenesis. This posits a rare population of oogonial stem cells that contribute new oocytes to partially compensate for the age-related decline in the primordial follicle reserve.
Jeffrey B Kerr, Michelle Myers and Richard A Anderson
Gillian Cowan, Andrew J Childs, Richard A Anderson and Philippa T K Saunders
The somatic (Sertoli cell (SC), Leydig cell (LC), and peritubular myoid (PTM) cell) cells play key roles in development of the fetal testis. We established monolayer cultures from second trimester human testes and investigated the pattern of expression of cell-lineage characteristic mRNAs. Expression of some SC-associated genes (SRY, SOX9, WT1, GATA4, and SF1) was detectable up to and including passage 3 (P3), while others (anti-Müllerian hormone; desert hedgehog) present prior to dissociation were not expressed in the cultured cells. Transcripts encoding the androgen receptor were expressed but addition of dihydrotestosterone (DHT) had no impact on expression of mRNAs expressed in SC or LC. Total concentrations of mRNAs encoding smooth muscle actin (ACTA2) and desmin increased from P1 to P3; an increasing proportion of the cells in the cultures were immunopositive for ACTA2 consistent with proliferation/differentiation of PTM cells. In conclusion, somatic cell monolayer cultures were established from human fetal testes; these cultures could form the basis for future studies based on isolation of purified populations of somatic cells and manipulation of gene expression that is difficult to achieve with organ culture systems. Our results suggest that fetal SC do not maintain a fully differentiated phenotype in vitro, yet PTM (ACTA2 positive) cells readily adapt to monolayer culture conditions in the presence of DHT. This culture system provides an opportunity to study the impact of regulatory factors on gene expression in PTM cells, a population thought to play a key role in mediating androgen action within the developing testis.
Roseanne Rosario, Hazel L Stewart, Emily Walshe and Richard A Anderson
In female mammals, reproductive potential is determined during fetal life by the formation of a non-renewable pool of primordial follicles. Initiation of meiosis is one of the defining features of germ cell differentiation and is well established to commence in response to retinoic acid. WIN 18,446 inhibits the conversion of retinol to retinoic acid, and therefore it was used to explore the impact of reduced retinoic acid synthesis on meiotic progression and thus germ cell development and subsequent primordial follicle formation. e13.5 mouse fetal ovaries were cultured in vitro and treated with WIN 18,446 for the first 3 days of a total of up to 12 days. Doses as low as 0.01 µM reduced transcript levels of the retinoic acid response genes Stra8 and Rarβ without affecting germ cell number. Higher doses resulted in germ cell loss, rescued with the addition of retinoic acid. WIN 18,446 significantly accelerated the progression of prophase I; this was seen as early as 48 h post treatment using meiotic chromosome spreads and was still evident after 12 days of culture using Tra98/Msy2 immunostaining. Furthermore, ovaries treated with WIN 18,446 at e13.5 but not at P0 had a higher proportion of growing follicles compared to vehicle controls, thus showing evidence of increased follicle activation. These data therefore indicate that retinoic acid is not necessary for meiotic progression but may have a role in the regulation of its progression and germ cell survival at that time and provide evidence for a link between meiotic arrest and follicle growth initiation.
Nicole A Bastian, Rosemary A Bayne, Katja Hummitzsch, Nicholas Hatzirodos, Wendy M Bonner, Monica D Hartanti, Helen F Irving-Rodgers, Richard A Anderson and Raymond J Rodgers
Fibrillins 1–3 are stromal extracellular matrix proteins that play important roles in regulating TGFβ activity, which stimulates fibroblasts to proliferate and synthesize collagen. In the developing ovary, the action of stroma is initially necessary for the formation of ovigerous cords and subsequently for the formation of follicles and the surface epithelium of the ovary. FBN3 is highly expressed only in early ovarian development and then it declines. In contrast, FBN1 and 2 are upregulated in later ovarian development. We examined the expression of FBN1–3 in bovine and human fetal ovaries. We used cell dispersion and monolayer culture, cell passaging and tissue culture. Cells were treated with growth factors, hormones or inhibitors to assess the regulation of expression of FBN1–3. When bovine fetal ovarian tissue was cultured, FBN3 expression declined significantly. Treatment with TGFβ-1 increased FBN1 and FBN2 expression in bovine fibroblasts, but did not affect FBN3 expression. Additionally, in cultures of human fetal ovarian fibroblasts (9–17weeks gestational age), the expression of FBN1 and FBN2 increased with passage, whereas FBN3 dramatically decreased. Treatment with activin A and a TGFβ family signaling inhibitor, SB431542, differentially regulated the expression of a range of modulators of TGFβ signaling and of other growth factors in cultured human fetal ovarian fibroblasts suggesting that TGFβ signaling is differentially involved in the regulation of ovarian fibroblasts. Additionally, since the changes in FBN1–3 expression that occur in vitro are those that occur with increasing gestational age in vivo, we suggest that the fetal ovarian fibroblasts mature in vitro.
Federica Lopes, Jin Liu, Stephanie Morgan, Rebecca Matthews, Lucy Nevin, Richard A Anderson and Norah Spears
Chemotherapy drugs are administered to patients using combination regimens, and as such the possibility of multiplicative effects between drugs need to be investigated. This study examines the individual and combined effects of the chemotherapy drugs cisplatin and doxorubicin on the human ovary. Although cisplatin and doxorubicin are known to affect female fertility, there is limited information about their direct effects on the human ovary, and none examining the possibility of combined, multiplicative effects of co-exposure to these drugs. Here, human ovarian biopsies were obtained from 14 women at the time of caesarean section, with 38 mouse ovaries also obtained from neonatal C57Bl/6J mice. Tissue was cultured for 6 days prior to analyses, with chemotherapy drugs added to culture medium on the second day of culture only. Treatment groups of a single (5 μg/mL human; 0.5 μg/mL mouse) or double (10 μg/mL human; 1.0 μg/mL mouse) dose of cisplatin, a single (1 μg/mL human; 0.05 μg/mL mouse) or double (2 μg/mL human; 0.01 μg/mL mouse) dose of doxorubicin or a combination of a single dose of both drugs together were compared to controls without drug exposure. Exposure to cisplatin or doxorubicin significantly decreased follicle health in human and mouse, supporting the suitability of the mouse as a model for the human ovary. There was also a significant reduction of mouse follicle number. Human ovarian stromal tissue exhibited increased apoptosis and decreased cell proliferation. Crucially, there was no evidence indicating the occurrence of multiplicative effects between cisplatin and doxorubicin.