The present results demonstrate (1) the presence of ACh in adult rat ovary and (2) indicate that GCs of growing antral ovarian follicles in both rodents and primates are the only detectable structures, which are immunoreactive for ChAT, implying that GCs are the source of ovarian ACh. Furthermore (3) our results indicate that ChAT staining increases in GCs of growing antral ovarian follicles in both rodents and primates and (4) show that FSH stimulates production of ACh in cells derived from an antral follicle. Finally, (5) since ChAT is absent from embryonic and neonatal ovary and embryonic development in ChAT-/- mice appears normal, these data support the hypothesis that ACh is a signaling molecule important for the antral follicle, which is produced by non neuronal cells of the ovary under gonadotrophin control.
A role of ACh in the ovary has only recently been suggested, based on studies with human GCs . These cells, derived from the preovulatory follicle of women undergoing in vitro fertilization procedures, bear resemblance to preovulatory follicular GCs. In the present study we sought to further define this ovarian site of presumed ACh synthesis.
We confirmed that in rodents and primates only growing follicles express ChAT, and thus, are likely to produce ACh. The ChAT protein is not found in embryonic and neonatal ovary and is first detected in follicles that are beginning to develop an antrum, i.e. a stage supported by FSH, which drives subsequent follicular growth [12, 13, 24]. Since preantral follicles lack ChAT, ACh may not be important for events taking place before antrum formation, including the embryonic phase of ovarian development. Unfortunately, ChAT null mice die at birth  precluding examination of postnatal ovarian development.
The increased ChAT immunoreactive signal detected in larger antral follicles, as compared with smaller antral follicles, suggested that ACh production may be regulated by FSH. This conclusion is supported by the results of our studies with immortalized GCs, which stably express functional FSH receptors , as these cells responded to FSH with increased production of ACh. This increase likely reflects an enhanced production of ACh per cell, because GFSHR-17 cells do not proliferate when treated with FSH , and the ACh values were expressed in relation to cellular protein content. A potential involvement of estrogens in ACh production can be ruled out (see ). Although estrogens are also produced by GCs of growing follicles in response to FSH, GFSHR-17 cells do not produce estrogens when challenged with FSH (see  for details). These cells however produce small amounts of progesterone . The cellular mechanisms employed by FSH to increase ACh production remain to be elucidated.
We clearly detected ACh itself by two techniques in homogenates from adult rat ovaries, in which ChAT immunoreactive GCs are present. Preliminary unpublished studies with immature rat ovarian homogenates (from ovaries of days 1–6) have sofar indicated that ACh in these samples is below detection limits. These results are consistent with the observed and documented lack of ChAT in the neonatal ovary (present study). More detailed studies, which are beyond the present report, are however required to examine ovarian ACh levels during sexual maturation and correlate them with ovarian and follicular development.
When measuring ovarian ACh content in adults we did not differentiate between ovaries bearing differently sized follicles or corpora lutea and the results obtained reflect a thus expected heterogeneity. Furthermore the two different methods employed provided somewhat different absolute values. Several reasons may account for these differences, which have to do with the principles of the assays used (HPLC detection versus enzymatic assay). Furthermore, it is possible that the magnitude of the changes is, at least in part, determined by changes in ACh metabolism. ACh is a very labile molecule with a very short half life time. Our preliminary data from fluorescence ACh assays strongly suggests presence of esterases degrading ACh, and preliminary Western blot data implicates butyrylcholine esterase [25, 26] as one of the esterases involved. Clearly, additional studies are required to elucidate the mechanisms of inactivation of ACh by GCs in the ovary.
ACh has only recently been recognized as a "cyto-transmitter" that, produced by non-neuronal cells, can act throughout the body in an autocrine and/or paracrine fashion (see ). It is now known that a variety of cells, including epithelial cells, endothelial cells, cancer cells, immune cells and placental tissue produce ACh and possess ACh receptors [16, 28, 29]. Importantly, ChAT activity of human placental cells is not affected by LH or FSH , suggesting that an FSH-dependent regulation in GCs is a cell-specific event.