Cholesterol transport and steroidogenesis by the corpus luteum

The synthesis of progesterone by the corpus luteum is essential for the establishment and maintenance of early pregnancy. Regulation of luteal steroidogenesis can be broken down into three major events; luteinization (i.e., conversion of an ovulatory follicle), luteal regression, and pregnancy induced luteal maintenance/rescue. While the factors that control these events and dictate the final steroid end products are widely varied among different species, the composition of the corpus luteum (luteinized thecal and granulosa cells) and the enzymes and proteins involved in the steroidogenic pathway are relatively similar among all species. The key factors involved in luteal steroidogenesis and several new exciting observations regarding regulation of luteal steroidogenic function are discussed in this review.     

Hormonal regulation and cell-specific expression of endothelin-converting enzyme 1 isoforms in bovine ovarian endothelial and steroidogenic cells

Endothelin-converting enzyme 1 (ECE-1) is a key enzyme in the biosynthesis of endothelin 1 (ET-1), a potent regulator of ovarian function. Different ECE-1 isoforms are localized in distinct intracellular compartments. Thus, the spatial and temporal pattern of ECE-1 expression determines the site of big ET-1 activation and the bioavailability of ET-1. This study was undertaken to investigate the hormonal regulation and cell-specific expression of ECE-1 isoforms in endothelial and steroidogenic cells of bovine follicles and corpora lutea (CL). Using enriched follicular and luteal cell subpopulations and in situ hybridization techniques, we showed that the ECE-1 gene is expressed by both endothelial and steroidogenic cells; however, the intracellular ECE-1a isoform was present only in ET-1-expressing endothelial cells. Steroidogenic cells in follicles or in CL, deficient in ET-1, expressed only the plasma membrane ECE-1b isoform. The intensity of antisense ECE-1 labeling in the granulosa cell layer increased with follicular size; insulin-like growth factor I and insulin upregulated ECE-1 expression when cultured with granulosa cells, suggesting that these growth factors may increase ECE-1 in growing follicles. In contrast, ET-1 and LH downregulated ECE-1 in steroidogenic cells. This effect could account for low ECE (and ET-1) levels, which characterize the early luteal phase. These findings suggest that ECE-1 is regulated during different stages of the cycle in a physiologically relevant manner. The hormonal regulation and intracellular localization of bovine ECE-1 isoforms revealed in this study may provide new insights into ET-1 biosynthesis and mode of action in different cellular microenvironments within the ovary.     

Proliferation of Luteal Steroidogenic Cells in Cattle

The rapid growth of the corpus luteum (CL) after ovulation is believed to be mainly due to an increase in the size of luteal cells (hypertrophy) rather than an increase in their number. However, the relationship between luteal growth and the proliferation of luteal steroidogenic cells (LSCs) is not fully understood. One goal of the present study was to determine whether LSCs proliferate during CL growth. A second goal was to determine whether luteinizing hormone (LH), which is known have roles in the proliferation and differentiation of follicular cells, also affects the proliferation of LSCs. Ki-67 (a cell proliferation marker) was expressed during the early, developing and mid luteal stages and some Ki-67-positive cells co-expressed HSD3B (a steroidogenic marker). DNA content in LSCs isolated from the developing CL increased much more rapidly (indicating rapid growth) than did DNA content in LSCs isolated from the mid CL. The cell cycle-progressive genes CCND2 (cyclin D2) and CCNE1 (cyclin E1) mRNA were expressed more strongly in the small luteal cells than in the large luteal cells. LH decreased the rate of increase of DNA in LSCs isolated from the mid luteal stage but not in LSCs from the developing stage. LH suppressed CCND2 expression in LSCs from the mid luteal stage but not from the developing luteal stage. Furthermore, LH receptor (LHCGR) mRNA expression was higher at the mid luteal stage than at the developing luteal stage. The overall results suggest that the growth of the bovine CL is due to not only hypertrophy of LSCs but also an increase in their number, and that the proliferative ability of luteal steroidogenic cells decreases between the developing and mid luteal stages.     

Histone acetyltransferase 1 (HAT1) acetylates hypoxia-inducible factor 2 alpha (HIF2A) to execute hypoxia response

Hypoxic response to low oxygen levels is characteristic of most solid cancers. Hypoxia-inducible factors (HIFs) regulate cellular metabolism, survival, proliferation, and cancer stem cell growth during hypoxia. The genome-wide analysis identified HAT1, a type B histone acetyltransferase, as an upregulated and essential gene in glioblastoma (GBM). GSEA analysis of differentially regulated genes in HAT1 silenced cells identified significant depletion of “hypoxia” gene sets. Hypoxia conditions induced HIF2A, not HIF1A protein levels in glioma cells in a HAT1-dependent manner. HAT1 and HIF2A interacted with each other and occupied the promoter of VEGFA, a bonafide HIF1A/HIF2A target. Acetylation of K512 and K596 residues by HAT1 is essential for HIF2A stabilization under normoxia and hypoxia as HIF2A carrying acetylation mimic mutations at either of these residues (H512Q or K596Q) showed stable expression in HAT1 silenced cells under normoxia and hypoxia conditions. Finally, we demonstrate that the HAT1-HIF2A axis is essential for hypoxia-promoted cancer stem cell maintenance and reprogramming. Thus, our study identifies that the HAT1-dependent acetylation of HIF2A is vital to executing the hypoxia-induced cell survival and cancer stem cell growth, therefore proposing the HAT1-HIF2A axis as a potential therapeutic target.     

Lipoprotein receptor expression during luteinization of the ovarian follicle

Ovarian follicles luteinize after ovulation, requiring structural and molecular remodeling along with exponential increases in steroidogenesis. Cholesterol substrates for luteal steroidogenesis are imported via scavenger receptor-BI (SR-BI) and the low-density lipoprotein (LDL) receptor from circulating high-density lipoproteins and LDL. SR-BI mRNA is expressed in pig ovaries at all stages of folliculogenesis and in the corpus luteum (CL). An 82-kDa form of SR-BI predominates throughout, is weakly present in granulosa cells, and is robustly expressed in the CL, along with the less abundant 57-kDa form. Digestion of N-linked carbohydrates substantially reduced the SR-BI mass in luteal cells, indicating that differences between forms is attributable to glycosylation. Immunohistochemistry revealed SR-BI to be concentrated in the cytoplasm of follicular granulosa cells, although found mostly at the periphery of luteal cells. To examine receptor dynamics during gonadotropin-induced luteinization, pigs were treated with an ovulatory stimulus, and ovaries were collected at intervals to ovulation. SR-BI in granulosa cell cytoplasm increased through the periovulatory period, with migration to the cell periphery as the CL matured. In vitro culture of follicles with human chorionic gonadotropin induced time-dependent upregulation of 82-kDa SR-BI in granulosa cells. SR-BI and LDL receptor were reciprocally expressed, with the latter highest in follicular granulosa cells, declining precipitously with CL formation. We conclude that luteinization causes upregulation of SR-BI expression, its posttranslational maturation by glycosylation, and insertion into luteal cell membranes. Expression of the LDL receptor is extinguished during luteinization, indicating dynamic regulation of cholesterol importation to maintain elevated steroid output by the CL.     

Size distribution of luteal cells from pregnant and non-pregnant marmoset monkeys and a comparison of the morphology of marmoset luteal cells with those from the human corpus luteum

The size distribution of marmoset luteal cells was determined on Days 6, 14 and 20 after ovulation in non-pregnant cycles and in early pregnancy. Image analysis was used to estimate the cell diameter of dispersed cells prepared from the marmoset corpus luteum (CL). Steroidogenic cells showed a size distribution consistent with one population of cells. There was a significant increase in mean cell diameter (P less than 0.05) from Day 6 to Day 14 in pregnant and non-pregnant animals with no further increase on Day 20. Micrographs of marmoset luteal tissue showed cells of greater than 10 micron containing the organelles typical of steroid-producing cells, and smaller non-steroidogenic cells surrounding the steroid-producing cells. On the basis of microscopy, there were no areas within the CL where cell composition was noticeably different. In contrast, micrographs of human luteal tissue showed two types of steroidogenic cell; most cells were similar to those in the marmoset CL but a smaller population of smaller cells could be distinguished around the periphery and along vascular septa. It is likely that these smaller and larger types of steroidogenic cells are of theca and granulosa cell origin respectively, the two cell populations differing in the degree of electron density and amount of rough endoplasmic reticulum. A distinguishing feature between marmoset and human luteal cells was the shape of the mitochondrian which were considerably rounder in marmoset luteal cells. The origin of steroidogenic cells in the marmoset CL is unclear, although in marmosets and man the luteal cell types display morphological characteristics distinct from the large and small luteal cells described for CL of the domestic ungulates.     

Growth factor modulation of steroidogenic acute regulatory protein and luteinization in the pig ovary.

In vivo and in vitro luteinization were investigated in the porcine ovary, with emphasis on expression of steroidogenic acute regulatory protein (StAR). StAR mRNA and protein as well as cytochrome P450 side-chain cleavage mRNA (P450scc) increased during the luteal phase in the corpus luteum (CL) and were absent in regressed CL. Cytochrome P450 aromatase mRNA (P450arom) was not detectable at any time in CL. In vitro luteinization of granulosa cells occurred over 96 h in culture, during which P450arom mRNA was present at 1 h after cell isolation but not detectable at 6 h; and P450scc and StAR mRNAs were first detectable at 6 h and 48 h, respectively. Incubation of cultures with insulin-like growth factor I (IGF-I, 10 ng/ml), dibutyryl cAMP (cAMP, 300 microM), or their combination, induced measurable StAR mRNA at 24 h (p < 0.05), increased progesterone accumulation at 48 h, and elevated both StAR and P450scc expression through 96 h. Incubation of luteinized granulosa cells with epidermal growth factor (EGF, 10 nM) changed their phenotype from epithelioid to fibroblastic, eliminated steady-state StAR expression, and interfered with cAMP induction of StAR mRNA and progesterone accumulation. EGF had little apparent effect on P450scc mRNA abundance. It is concluded that StAR expression characterizes luteinization, and early luteinization is induced by cAMP and IGF-I in vitro. Further, EGF induces a morphological and functional phenotype that appears similar to an earlier stage of granulosa cell function.     

In vitro differentiation of bovine theca and granulosa cells into small and large luteal-like cells: morphological and functional characteristics

This study was undertaken to investigate whether bovine granulosa and theca interna cells could be luteinized in vitro into luteal-like cells. Granulosa and theca cells were cultured for 9 days in the presence of forskolin (10 microM), insulin (2 micrograms/ml), insulin-like growth factor I (100 ng/ml), or a combination of these agents. During the first day of culture, granulosa and theca cells secreted estradiol and androstenedione, respectively; progesterone rose only after 3-5 days in culture and reached a maximum on the ninth day of culture. Cells incubated in the presence of forskolin plus insulin exhibited morphological and functional characteristics of luteal cells isolated from the corpus luteum. It was found that cell diameter, basal and stimulated progesterone secretion, and pattern of cell replication for both cell types were comparable to those of luteal cells. Numerous lipid droplets and intensified mitochondrial adrenodoxin staining also indicated active steroidogenesis in luteinized cells. After 9 days in culture, stimulants were withdrawn, and the culture proceeded in basal medium for an additional 5 days; elevated progesterone levels were maintained by luteinized granulosa cells (LGC), whereas in contrast a dramatic drop in progesterone production was observed in luteinized theca cells (LTC). On Day 9, cells were challenged for 3 h with LH (10 ng/ml), forskolin (10 microM), or cholera toxin (100 ng/ml), resulting in a 4-fold increase in progesterone secretion by LTC; the same treatments failed to stimulate progesterone in LGC.(ABSTRACT TRUNCATED AT 250 WORDS)     

Influence of high-density lipoprotein on estradiol stimulation of luteal steroidogenesis

The aim of this investigation was to determine whether luteal cells utilize cholesterol derived from high-density lipoprotein (HDL) for steroidogenesis and whether estrogen enhances luteal utilization of exogenous sterol. Incubation of Day 15 corpora lutea (CL) with different doses of human HDL resulted in a dose-dependent increase in progesterone production. HDL in vitro enhanced the overall steroidogenic capacity. However, the percentage of increases in 17 alpha-hydroxyprogesterone, testosterone and estradiol were significantly less than that of progesterone. Day 12 hypophysectomized and hysterectomized pregnant rats were treated with either estradiol, testosterone or vehicle for 72 h. Serum pregnenolone and progesterone were markedly increased by the steroid treatment, yet in vitro production of progesterone by CL in all the groups was similar. However, in the presence of HDL in the media, only luteal tissues from steroid-treated rats increased their progesterone output. The reduced production of progesterone by luteal cells of vehicle-treated rats was not due to an accumulation of pregnenolone but to an overall reduction in exogenous sterol utilization. In summary, results of this investigation suggest 1) luteal cells of pregnant rats effectively utilize cholesterol from HDL for maximal steroidogenesis, and 2) estradiol may stimulate luteal steroidogenesis, at least in part, by affecting the incorporation or utilization of cholesterol from HDL into the cell.     

The role of autophagy in corpus luteum regression in the rat

Autophagy is associated with luteal cells death during regression of the corpus luteum (CL) in some species. However, the involvement of autophagy or the association between autophagy and apoptosis in CL regression are largely unknown. Therefore, we investigated the role of autophagy in CL regression and its association with apoptosis. Ovaries were obtained from pseudopregnant rats at Days 2 (early), 7 (mid-), and 14 and 20 (late-luteal stage) of the pseudopregnancy; autophagy-associated protein (microtuble-associated protein light chain 3 [LC3]) was immunolocalized and its expression level was measured. Luteal cell apoptosis was evaluated by measuring cleaved caspase 3 expression. LC3 expression increased slightly from early to mid-luteal stage, with maximal levels detected at the late-luteal stage in steroidogenic luteal cells. The expression level of the membrane form of LC3 (LC3-II) also increased during luteal stage progression, and reached a maximum at the end point of late-luteal stage (Day 20). This pattern coincided with cleaved caspase 3 expression. Furthermore, LC3-II expression increased, as did levels of cleaved caspase 3 in luteal cells cultured with prostaglandin F(2alpha) known to induce CL regression. These findings suggest that luteal cell autophagy is directly involved in CL regression, and is correlated with increased apoptosis. In addition, autophagic processes were inhibited using 3-methyladenine or bafilomycin A1 to evaluate the role of autophagy in apoptosis induction. Inhibition of autophagosome degradation by fusion with lysosomes (bafilomycin A1) increased apoptosis and cell death. Furthermore, inhibition of autophagosome formation (3-methyladenine) decreased apoptosis and cell death, suggesting that the accumulation of autophagosomes induces luteal cell apoptosis. In conclusion, these results indicate that autophagy is involved in rat luteal cell death through apoptosis, and is most prominent during CL regression.     

Requirement of the Fas ligand-expressing luteal immune cells for regression of corpus luteum

Apoptosis in corpus luteum (CL) is induced by prolactin (PRL) in female rats. PRL-induced apoptosis in CL is mediated by the Fas/Fas ligand (FasL) system. The CL consists of steroidogenic and non-steroidogenic cells, including immunocytes. Fas mRNA was detected only in the luteal steroidogenic cells, and FasL mRNA was expressed only by the non-steroidogenic CD3-positive luteal immunocytes. Removing the luteal immune cells from the luteal cells inhibited PRL-induced luteal cell apoptosis effectively. Thus, FasL-expressing non-steroidogenic luteal immunocytes are required for PRL-induced luteal cell apoptosis and heterogeneous induction of apoptosis by Fas/FasL in CL.     

Expression and localisation of ephrin-B1 and EphB4 in steroidogenic cells in the naturally cycling mouse ovary

Ephrin receptors and ligands are membrane-bound molecules that modulate diverse cellular functions such as cell adhesion, epithelial–mesenchymal transition, motility, differentiation and proliferation. We recently reported the co-expression of ephrin-B1 and EphB4 in adult and foetal Leydig cells of the mouse testis, and thus speculated that their co-expression is a common property in gonadal steroidogenic cells. Therefore, in this study we examined the expression and localisation of ephrin-B1 and EphB4 in the naturally cycling mouse ovary, as their expression patterns in the ovary are virtually unknown. We found that ephrin-B1 and EphB4 were co-expressed in steroidogenic cells of all kinds, i.e. granulosa cells and CYP17A1-positive steroidogenic theca cells as well as in 3β-HSD-positive luteal cells and the interstitial glands; their co-expression potentially serves as a good marker to identify sex steroid-producing cells even in extra-gonadal organs/tissues. We also found that ephrin-B1 and EphB4 expression in granulosa cells was faint and strong, respectively; ephrin-B1 expression in luteal cells was weak in developing and temporally mature corpora lutea (those of the current cycle) and likely strong in regressing corpora lutea (those of the previous cycle) and EphB4 expression in luteal cells was weak in corpora lutea of the current cycle and likely faint/negative in the corpora lutea of the previous cycle. These findings suggest that a luteinising hormone surge triggers the upregulation of ephrin-B1 and downregulation of EphB4, as this expression fluctuation occurs after the surge. Overall, ephrin-B1 and EphB4 expression patterns may represent benchmarks for steroidogenic cells in the ovary.     

The expression of prolactin and its cathepsin D-mediated cleavage in the bovine corpus luteum vary with the estrous cycle

In the corpus luteum (CL), blood vessels develop, stabilize, and regress. This process depends on the ratio of pro- and antiangiogenic factors, which change during the ovarian cycle. The present study focuses on the possible roles of 23,000 (23K) prolactin (PRL) in the bovine CL and its antiangiogenic NH(2)-terminal fragments after extracellular cleavage by cathepsin D (Cath D). PRL RNA and protein were demonstrated in the CL tissue, in luteal endothelial cells, and in steroidogenic cells. Cath D was detected in CL tissue, cell extracts, and corresponding cell supernatants. In the intact CL, 23K PRL levels decreased gradually, whereas Cath D levels concomitantly increased between early and late luteal stages. In vitro, PRL cleavage occurred in the presence of acidified homogenates of CL tissue, cells, and corresponding cell supernatants. Similar fragments were obtained with purified Cath D, and their appearance was inhibited by pepstatin A. The aspartic protease specific substrate MOCAc-GKPILF~FRLK(Dnp)-D-R-NH(2) was cleaved by CL cell supernatants, providing further evidence for Cath D activity. The 16,000 PRL inhibited proliferation of luteal endothelial cells accompanied by an increase in cleaved caspase-3. In conclusion, 1) the bovine CL is able to produce PRL and to process it into antiangiogenic fragments by Cath D activity and 2) PRL cleavage might mediate angioregression during luteolysis.