In vitro differentiation of theca cells from ovarian cells isolated from postmenopausal women.

<bold>Study Question: </bold>Can human theca cells (TCs) be differentiated in vitro?<bold>Summary Answer: </bold>It is possible to differentiate human TCs in vitro using a medium supplemented with growth factors and hormones.<bold>What Is Known Already: </bold>There are very few studies on the origin of TCs in mammalian ovaries. Precursor TCs have been described in neonatal mice ovaries, which can differentiate into TCs under the influence of factors from oocytes and granulosa cells (GCs). On the other hand, studies in large animal models have reported that stromal cells (SCs) isolated from the cortical ovarian layer can also differentiate into TCs.<bold>Study Design, Size, Duration: </bold>After obtaining informed consent, ovarian biopsies were taken from eight menopausal women (53-74 years of age) undergoing laparoscopic surgery for gynecologic disease not related to the ovaries. SCs were isolated from the ovarian cortex and in vitro cultured for 8 days in basic medium (BM) (G1), enriched with growth factors, FSH and LH in plastic (G2) or collagen substrate without (G3) or with (G4) a GC line.<bold>Participants/materials, Setting, Methods: </bold>To confirm TC differentiation, relative mRNA levels for LH receptor (Lhr), steroidogenic acute regulatory protein (Star), cholesterol side-chain cleavage enzyme (Cyp11a1), cytochrome P450 17A1 (Cyp17a1), hydroxy-delta-5-steroid dehydrogenase, 3 beta- and steroid delta-isomerase 1 (Hsd3b1) and hydroxy-delta-5-steroid dehydrogenase, 3 beta- and steroid delta-isomerase 2 (Hsd3b2) were assessed. Immunohistochemistry was also performed for their protein detection and a specific marker was identified for TCs (aminopeptidase-N, CD13), as were markers for theca and small luteal cells (dipeptidyl peptidase IV (CD26) and Notch homolog 1, translocation-associated (NOTCH1)). Finally, we analyzed cell ultrastructure before (Day 0) and after in vitro culture (Day 8), and dehydroepiandrosterone (DHEA) and progesterone levels in the medium using transmission electron microscopy (TEM) and ELISA, respectively.<bold>Main Results and the Role Of Chance: </bold>Results obtained from qPCR showed a significant increase (P < 0.05) in mRNA levels of Lhr in F2 (floating cells in G2) and G4, Cyp17a1 in G1 and F1 (floating cells in G1) and Hsd3b2 in G1, G2, G3 and G4. Immunohistochemistry confirmed expression of each enzyme involved in the steroidogenic pathway at the protein stage. However, apart from G1, all other groups exhibited a significant (P < 0.05) rise in the number of CD13-positive cells. There was also a significant increase (P < 0.05) in NOTCH1-positive cells in G3 and G4. Ultrastructure analyses by TEM showed a distinct difference between groups and also versus Day 0. A linear trend with time revealed a significant gain (q < 0.001) in DHEA concentrations in the medium during the culture period in G1, G2, G3 and G4. It also demonstrated a statistical increase (q < 0.001) in G2, G3 and G4 groups, but G1 remained the same throughout culture in terms of progesterone levels.<bold>Large Scale Data: </bold>N/A.<bold>Limitations, Reasons For Caution: </bold>Shorter periods of in vitro culture (e.g. 2, 4 and 6 days) could have led to increased concentrations of differentiated TCs in G2, G3 and G4. In addition, a group of cells cultured in BM and accompanied by COV434 cells would be necessary to understand their role in the differentiation process. Finally, while our results demonstrate that TCs can be differentiated in vitro from cells isolated from the cortical layer of postmenopausal ovaries, we do not know if these cells are differentiated from a subpopulation of precursor TCs present in ovarian cortex or ovarian SCs in general. It is therefore necessary to identify specific markers for precursor TCs in human ovaries to understand the origin of these cells.<bold>Wider Implications Of the Findings: </bold>This is a promising step toward understanding TC ontogenesis in the human ovary. Moreover, in vitro-generated human TCs can be used for studies on drug screening, as well as to understand TC-associated pathologies, such as androgen-secreting tumors and polycystic ovary syndrome.<bold>Study Funding/competing Interest(s): </bold>This study was supported by grants from the Fonds National de la Recherche Scientifique de Belgique (FNRS) (C.A.A. is an FRS-FNRS Research Associate; grant MIS #F4535 16 awarded to C.A.A.; grant 5/4/150/5 awarded to M.M.D.; grant ASP-RE314 awarded to P.A.) and Foundation Against Cancer (grant 2018-042 awarded to A.C.). The authors declare no competing interests. [ABSTRACT FROM AUTHOR]