Studies on the regulation of progesterone biosynthesis in the human placenta

Abstract

The placenta is a multifunctional, ephemeral organ that develops during mammalian pregnancy as an association between fetal and maternal tissues. It performs a wide variety of functions to protect and nourish the fetus until it can exist independently. The placenta physically anchors the fetus to the uterus; transports nutrients from the maternal circulation to the fetus; excretes fetal wastes into the maternal compartment; and plays an immunomodulatory role in maternal acceptance of the fetal semiallograft. Apart from these functions, it serves as an extremely efficient endocrine organ (well exemplified by the human placenta) by elaborating several protein, peptide and steroid hormones that regulate maternal and fetal physiology. Among these, the most important hormone for maintenance of pregnancy is progesterone. Progesterone is absolutely indispensable throughout gestation, and its absence or inhibition can lead to termination of pregnancy. Its primary role is to inhibit uterine smoothmuscle contraction and prevent cervical dilation, thus preventing expulsion of the fetus. Progesterone is also postulated to have an immunosuppressive role during pregnancy. Additionally, progesterone stimulates synthesis and secretion of human chorionic gonadotropin (hCG), a key hormone in early pregnancy maintenance. Despite its essential role, very little is known about the factors regulating progesterone biosynthesis. The human placenta acquires the capacity to synthesize progesterone only around the ninth week of pregnancy. Prior to this, the corpus luteum of the fertile cycle produces the progesterone required to maintain pregnancy. hCG, produced by trophoblasts under progesterone influence, rescues the corpus luteum from degeneration and maintains luteal progesterone production until the placenta becomes steroidogenically competent. The human placenta produces large amounts of progesterone during pregnancy-up to 250-600 mg per day in late gestation. However, hCG has no stimulatory effect on placental progesterone production. Further, the human placenta cannot synthesize cholesterol, the precursor for progesterone, de novo from acetate. It depends entirely on maternal cholesterol supplied as lowdensity lipoprotein (LDL). Uptake of LDL is one of the ratelimiting steps in placental progesterone biosynthesis and is mediated by LDL receptors on the brushborder membrane of placental villi. Paradoxically, LDLreceptor numbers decline in late gestation-a period of maximal progesterone production. Since placental steroidogenesis unlike other steroidogenic tissues (testis, ovary, adrenal cortex) does not appear to be driven by trophic hormones, a strong intrinsic regulator is likely required to maintain high progesterone output. Thus, placental steroids-estradiol and progesterone-may themselves regulate placental progesterone synthesis. The present study investigates the roles of progesterone and estradiol in regulating placental progesterone biosynthesis.

Chapter 1 This chapter provides a brief account of the development, structure and functions of the human placenta, followed by a detailed discussion of progesterone synthesis, its regulation, and the mechanism of its action via progesterone receptors in nonplacental tissues.

Progesterone Autoregulation and Identification of Placental Progesterone Receptors Because no established regulator of placental progesterone synthesis is known, progesterone production might be under autoregulatory control. To test this hypothesis, it is essential first to establish that the human placenta is a target tissue for progesterone by demonstrating the presence of progesterone receptors (PR). Reports in the literature are contradictory regarding the presence or absence of PR in the human placenta. RTPCR, a sensitive method capable of detecting even very low mRNA levels, was therefore used. Primers targeted to the ligandbinding domain of the human PR were used to analyse RNA isolated from firsttrimester human placenta (FTHP) and term placenta. A 351 bp DNA fragment corresponding to PR was amplified from both FTHP and term placenta. Southern blotting using either fulllength PR cDNA or the PR ligandbinding domain confirmed that the 351 bp product specifically hybridized to PR probes. Since RTPCR is extremely sensitive and could amplify transcripts from trace uterine contamination (uterus is closely associated with placenta and rich in PR), it was necessary to rule out contamination. Expression of a uterusspecific gene, PP14, was checked in the same RNA samples. PP14 was not amplified, although PR was, confirming the placental origin of the PR transcript. Further validation came from restriction mapping of the 351 bp product using three enzymes, which produced the expected digestion pattern. The fragment was cloned into pRSETA, and two independent clones were sequenced. Their sequences matched published human PR cDNA, confirming the presence of PR in the human placenta. These results are described in Chapter 2. The fact that PR mRNA is detectable only by RTPCR and not by Northern blot analysis indicates that the levels of PR mRNA are low in the human placenta. As a result, the physiological relevance of the mRNA present in such low quantities becomes questionable. In view of this, it was essential to demonstrate the presence of PR protein in the human placenta and to establish that it is functional. The PR belongs to the steroid/thyroid/retinoid receptor superfamily of molecules that function as liganddependent transcription factors. Since it is a DNAbinding protein, gelretardation assays can be conveniently used to demonstrate the presence of a functional PR protein. Oligonucleotides corresponding to the progestin response element (PRE) were synthesized, and the duplex PRE was radiolabeled and used for binding studies. When nuclear extracts from FTHP or term placenta were incubated with radiolabeled PRE, three retarded complexes were observed, designated arbitrarily as complexes I, II, and III. All three complexes were specific, since their formation could be abolished by prior incubation with nonradioactive PRE but not nonspecific DNA. To identify which complex contained PR, antibody supershift assays were performed. Monoclonal antibodies (MAb) specific to human PR were obtained from Dr. G. L. Greene (BenMay Institute, University of Chicago) and Dr. P. G. Satyaswaroop (University of Pennsylvania). Using hPRa3, a MAb that recognizes both PR isoforms (PRA and PRB), it was observed that only complex I contained PR. Complexes II and III may contain other members of the steroidreceptor superfamily capable of binding PRE. Westernblot analysis using hPRa3 detected two signals corresponding to Mr 82 kDa and 68 kDa in both FTHP and term placenta. The 82 kDa species corresponds to the reported size of PRA. The 68 kDa species is likely a proteolytic product, previously observed in other tissues. The reported Mr of PRB (116-120 kDa) was not detected in either FTHP or term placenta. To confirm this, supershift assays using KC146 (a PRBspecific MAb) were performed. KC146 failed to supershift complex I in placental extracts but did so in T47D breastcancer nuclear extracts (known to express PRB). This is the first report showing expression of only PRA in any mammalian tissue; all other PRexpressing cell types contain both PRA and PRB. These results are described in Chapter 3.

Having demonstrated the presence of functional PR in the human placenta, the role of progesterone in regulating placental progesterone biosynthesis was examined. Because this required measuring progesterone levels in placental mince preparations incubated with different modulators, a sensitive radioimmunoassay (RIA) for progesterone was developed. Direct addition of progesterone could not be tested because exogenous progesterone would interfere with the measurement. To circumvent this limitation, specific antiprogestins-RU 486 and ZK 98299-were used. These antagonize endogenous progesterone at the receptor level, mimicking effects opposite to progesterone. Neither RU 486 nor ZK 98299 crossreacted with the progesterone antiserum used for RIA, allowing their use without assay interference. RU 486 and ZK 98299 showed paradoxical effects: both stimulated progesterone production in FTHP but inhibited production in term placenta. Although this appears to imply negative autoregulation, this contradicts the wellestablished requirement for high progesterone throughout gestation. It is known that RU 486 and ZK 98299 can act as agonists under certain conditions. Their effects depend on several factors, including:

the PR isoform involved phosphorylation state of PR coactivators and corepressors crosstalk with other signaling pathways

Collectively, these form the “functional unit” of PR. To explain the paradox, the sensitivity of PREbinding complexes to ZK 98299 was examined using gelretardation assays. ZK 98299, a class II antiprogestin, reduces PR affinity for PRE. In FTHP, ZK 98299 inhibited formation of only complex I (the PRcontaining complex), whereas in term placenta all three complexes were inhibited. Because complexes II and III also bind PRE, they can compete with PR for binding and therefore modulate PR function. Their differential sensitivity suggests that the PR functional unit differs between FTHP and term placenta, potentially explaining the paradoxical antiprogestin effects. Although these experiments suggest that progesterone regulates its own synthesis, they do not establish whether progesterone stimulates or inhibits it. To address this, the effect of exogenous progesterone on 3hydroxysteroid dehydrogenase (3HSD) mRNA levels was studied. Progesterone increased 3HSD mRNA, indicating that progesterone stimulates its own synthesis in the placenta. Estradiol, another major placental steroid, also plays a key role in pregnancy maintenance. Estradiol is known to potentiate progesterone action by inducing PR synthesis. Addition of estradiol to FTHP or term placenta increased progesterone synthesis significantly. This was confirmed through inhibition experiments using:

CGS 16949A (aromatase inhibitor) ICI 182780 (estrogen receptor antagonist)

both of which reduced progesterone synthesis. To examine the mechanism, estradiol’s effect on LDLreceptor mRNA was studied. Estradiol increased LDLreceptor mRNA levels, suggesting that elevated LDLreceptor expression increases LDL uptake-the ratelimiting substrate for progesterone synthesis. These results are presented in Chapter 4, and Chapter 5 provides a general discussion.