Linda Tseng

Estrogen sulfatase and estrogen sulfotransferase in human primary mammary carcinoma

Estrogen sulfatase (ES) and estrogen sulfotransferase (ESFT) activities were measured in a group of primary breast tumors. The mean value of ES activities, measured in 66 breast tumor specimens, was 0.9 nmol of estrone formed from estrone sulfate/mg tissue protein per hr regardless of the hormone receptor status of the specimen. However, the average value of the ESFT activity, expressed in nmol of estradiol-3-sulfate (E2S) formed from estradiol (E2)/mg of cytosol protein per hr, was found to be significantly higher in ER +/PGR + tumors (n = 26, 0.18 +/- 0.15, means +/- SD) than in ER -/PGR - tumors (n = 31, 0.08 +/- 0.06, P less than 0.005). Normal breast tissues also contain ES and ESFT but the activities were lower than those in tumors. When fresh breast tumor tissue fragments were incubated with radioactive E2 (0.4 nM) and E2S (3 nM) separately, E2 was not sulfurylated appreciably while E2S was extensively hydrolyzed to free estrogens indicating that the combined effect of ES and ESFT in breast tumor is favored towards the hydrolysis of estrogen sulfate. These results imply that the circulating estrogen sulfate could be utilized as the precursor of active estrogen to promote the cell growth in hormone sensitive tumors.

Ligand activated hPR modulates the glycodelin promoter activity through the Sp1 sites in human endometrial adenocarcinoma cells.

Human endometrium produces glycodelin-A (GdA). The GdA mRNA is highly expressed in progestin-sensitized human endometrial glandular epithelial cells. The mechanism of GdA gene expression, however, is not clear. To understand the cell specific GdA gene transcription, our first approach was to identify the cis-element in the GdA promoter using transfection assay in a human endometrial adenocarcinoma cell line (HEC-1B, a cell line originally derived from the glandular component of the endometrium). The GdA promoter (-1900 to +20 bp) was linked to the luciferase reporter gene to construct p1900Luc, along with two shorter promoter constructs, p1100Luc and p304Luc. Deletion analysis showed that the basal promoter activity was derived from the region between -304 to +20 bp. This region contains three putative Sp1 binding sites (Sp1-1, -243 to -238 bp; Sp1-2, -207 to -202 bp; and Sp1-3, -56 to -49 bp). Mutation analysis at the Sp1 sites showed that p304Spm2Luc and p304Spm3Luc reduced the activity by 80%, while p304Spm1-2-3Luc reduced the activity by 95%. Sp1-1 mutation, however, had no effect. These results showed that two of the three Sp1 cis-elements mediate the basal promoter activity of the GdA gene. Electrophoretic gel mobility shift showed that at least two specific binding proteins in the nuclear extracts of HEC-1B cells bound to the oligo containing Sp1-2 or Sp1-3 cis-element. Sp1 antibody reduced the specific binding complex by 70% suggesting that Sp1 transcription factor regulates GdA gene expression. In addition, over expression of Sp1 increased the promoter activity. To determine whether progestin would modulate the promoter activity, HEC-1B cells were transfected with p304Luc and with progesterone receptor (either hPR-A or hPR-B) expression vector. Medroxyprogesterone acetate increased the promoter activity (3-fold) derived from p304Luc but not from the mutant, p304Spm1-2-3Luc. In contrast, the promoter activity was slightly reduced in cells treated with estradiol and co-transfected with estrogen receptor expression vector. These data indicate that ligand-activated PR stimulates GdA gene expression mediated through the functional Sp1 sites.

Progesterone Receptor (hPR) Upregulates the Fibronectin Promoter Activity in Human Decidual Fibroblasts

Previous studies have shown that progestin induces the production of fibronectin (FN) and its mRNA content in human endometrial stromal cells. The mechanism of the upregulation was unclear. In the present study, we provide evidence that hPR regulates the FN promoter activity mainly through the CRE/AP1 site located in the proximal region of the promoter in human decidual fibroblasts. Various lengths of the proximal region of the FN promoter were linked to the reporter vector to construct promoter-reporter plasmids and were then transfected into human decidual fibroblasts. Deletion and mutation analysis showed that CRE/AP1 and Sp1 sites in the proximal region mediated the basal promoter activity. To evaluate progestin-mediated transcriptional activation, decidual fibroblasts were transfected with p300 (FN promoter-reporter construct) and hPR expression vector. Cells treated with medroxyprogesterone acetate (MPA) increased the promoter activity ranging from 2.5- to 9-fold determined in 10 decidual specimens. hPRA enhanced activation was stronger than that of hPRB. Structural analysis of hPR showed that DNA and ligand binding domains are essential for the activation, and missing the TAF1 domain weakens the activation. The proximal promoter region of the FN gene lacks a canonical PRE site. Mutation at the CRE/AP1 site eliminated the upregulation by progestin. To evaluate the interaction of hPR with the CRE/AP1 site, the CRE/AP1 site was mutated to the consensus AP1 cis-element (TGACGTCA, -172 to -165 bp, mutated to TGAC_TCA) which eliminated the CREB binding. FN promoter activity derived from p300AP1 mutant was found to be higher than that of p300. These results showed that hPR interacts with the AP1 binding proteins, but not with CREB. Progestin treatment or overexpression hPR did not alter appreciably the content of c-jun or c-fos in decidual fibroblasts nuclear extracts. Antibody to hPR (hPRa3), which precipitated hPR also coprecipitated c-jun and c-fos, whereas CREB was not precipitated by hPRa3. The observation implies that hPRs are brought to the FN promoter region by AP1 proteins to enhance the transcription. In summary, this study provides molecular evidence that the CRE/AP1 site and c-jun/c-fos in decidual fibroblasts mediate the hPR-enhanced activation of FN transcription.

Kinetic studies of human endometrial hydroxysteroid dehydrogenase

Abstract Human endometrial hydroxysteroid dehydrogenase (HSD) in microsomal and mitochondrial fractions was characterized by the kinetic analysis of eight steroids: estradiol† (E2), estrone (E1), testosterone (T), androstenedione (A), 5-ene-androstene-3β,17β-diol (Adiol), dehydroepiandrosterone (D), 20α-dihydroprogesterone (20α-DHP) and progesterone (P). The enzyme activity was estimated from the initial rate of the formation of the product in the presence of various concentrations of [3H]-steroid and an excess amount of cofactor NAD for the oxidation and NADH for the reduction. The oxidation of E2 had a smaller km and a larger vmax than the reduction of E1, indicating that the enzyme favors oxidation. This observation correlates with the previous findings that in the intact cell the interconversion between E2 and E1 is favored toward the formation of E1. The effect of various steroids on the rate of oxidation of E2 and 20α-DHP was examined kinetically. The reaction was found to be competitively inhibited, suggesting that various HSD activities may derive from a single enzyme. The inhibitor constant, kiK, is similar to the km value for each steroid. The relative ki values indicate that E2 has the highest affinity to the enzyme followed in order by 20α-DHP, Adiol, T and A. The endometrial HSD in particulate fractions was treated with trypsin, phospholipase A2 and detergents. The HSD activity was significantly lost after incubation with these enzymes, especially with phospholipase A2, and detergents, indicating that HSD is tightly associated with the particulate components and the activity is stabilized by binding to the membrane.

Progesterone receptor activates its promoter activity in human endometrial stromal cells

Previous studies have shown that progestin increases the content of progesterone receptor (hPRA and hPRB) and the hPR mRNA during decidualization of human endometrial stromal cells suggesting that endogenous hPR enhances the transcription of the hPR gene. In the present study, we provide evidence that hPR regulates the promoter activity mediated through an active Sp1 site. In stromal cells treated with medroxyprogesterone acetate, the promoter activity was significantly increased when cells were co-transfected with hPR expression vector. Progressive deletion analysis showed that the highest activity was derived from the promoter region between -55 and +31 bp. Transactivation by hPR was dose dependent. The capacity of hPRA was stronger than that of hPRB. The ligand binding domain, but not DNA binding domain of the hPR was required for the transactivation. The proximal promoter region lacks a canonical progesterone response element. Instead, an active Sp1 site (-49 to -43 bp) has been confirmed. Mutation of the Sp1 site eliminated the effect of hPR activation. The promoter activity was increased by over expression of Sp1, whereas Sp3 had no effect. Electrophoretic mobility shift assay showed that the promoter region between -55 and +31 bp bound to Sp1 family proteins, Sp1 (C2 complex) and Sp3 (C1 and C3 complexes) identified by antibodies to Sp1 and Sp3. Sp1 complex formed by extracts of stromal cells was less intense than that formed by progestin-decidualized stromal cells. Sp1/DNA binding was enhanced when stromal cell extracts were incubated with calf intestine alkaline phosphatase (CIP) suggesting that dephosphorylation of Sp1 enhances the DNA binding. Addition of protein kinase inhibitor, H-89 or H-7, enhanced the hPR stimulated promoter activity. Western blot analysis showed that endometrial stromal/decidual cell extracts contained a wide band of Sp1 spanning from approximately 105 to 96 kDa and was resolved into one band at 96 kDa by CIP. Decidual cell extracts are abundant with the 96 kDa Sp1. In addition, the 96 kDa Sp1 was co-precipitated with ligand-activated hPRA or hPRB in the decidual cell nuclear extracts. These data suggest that dephosphorylated Sp1, abundant in decidual cells, enhances the binding to both DNA and hPR resulting in a robust increase of the hPR promoter activity.

Estrogen synthesis in human endometrial epithelial glands and stromal cells

Biosynthesis of estrogen was studied in human endometrial epithelial glands and stromal cells. Intact epithelial glands and stromal cells were isolated from human normal proliferative and secretory endometrial specimens. The isolated glands and stromal cells were separately incubated with 10 nM [3H]testosterone in nutrient medium for 20 h in an incubator. Only estrone and estradiol were formed in stromal cells and in epithelial glands in the proliferative phase. Four types of estrogens: estrone, estradiol, estrone sulfate and estradiol-3-sulfate were isolated from the secretory epithelial glands. The capacity of aromatization is higher in stromal cells than that in glands. The highest capacity of aromatization was found in stromal cells in the proliferative phase, 58 +/- 30 fmol/mg protein (means +/- SD, n = 10) followed in order by stromal cells in the secretory phase, 30 +/- 18 (n = 6), and the epithelial glands in the proliferative and secretory phase, 22 +/- 12 (n = 10) and 4.3 +/- 1.2 (n = 6), respectively. These results indicate that the aromatase in endometrium is distributed in both stromal cells and epithelial glands. The capacity of aromatization varies in the 2 phases of the menstrual cycle with a significant decrease in activity of the epithelial glands in the secretory phase. The capacity of aromatization was also examined when stromal cells were incubated with various concentrations of [3H]T. The apparent Km, estimated from the concentration of delta 4-androgens in cells, is about 27 nM, indicating that the endometrial aromatase has a considerably high affinity to delta 4-androgens.

Cyclic Change of Estradiol Metabolic Enzymes in Human Endometrium During the Menstrual Cycle

Estradiol dehydrogenase (E2DH), estrogen sulfatase and sulfotransferase are the three major estrogen metabolic enzymes found to be present in the human endometrium. The activities of these enzymes were compared in a group of normal endometrial specimens at various stages of the menstrual cycle. The average value of E2DH activity estimated from the rate of conversion of tritiated estradiol (E2) to estrone (E1), in the presence of an excess amount of NAD, was 2.8 nmol of El formed/mg protein/h in proliferative endometrium and 22 in secretory endometrium. The estrogen sulfatase activity was estimated by measuring the rate of conversion of [3H]-estrone sulfate (E1S) to El. The average value was 1.6 nmol/mg protein/h with no significant change in activity during the menstrual cycle. The sulfotrnsferase activity, estimated from the rate of conversion of [3H]-E1 to [3H] -E1S, in the presence of an excess amount of adenosine 3’-phosphate 5-phosphosulfate (PAPS) had an average value of 0.03 nmol/mg protein/h in the secretory endometrium but was not detectable in the proliferative phase.

Notch ligand-dependent gene expression in human endometrial stromal cells.

The purpose of this study is to investigate the expression patterns and role of Notch signaling in human endometrial cells. Notch receptors, Notch 1-3 were expressed in both endometrial epithelial and stromal cells. Notch ligands, Jag1 and Dll4 and Notch target genes, Hes1 and Hey1 were predominantly expressed in endometrial epithelial cells and scarce in stromal cells. Increased de novo synthesis of Dll4 or Jag1 in stromal cells by retroviral delivery significantly induced Hes1 and Hey1. Evaluations of global gene expression by microarrays revealed that more than 400 genes in stromal cells were significantly regulated by Jag1. Gene annotation-based functional analysis classified these genes into biological processes of cell adhesion, cell structure and motility, cell communication, cell cycle, and angiogenesis. This study provides evidence that Notch ligands control the Notch gene activities and may enhance development of human endometrium.

Hox proteins activate the IGFBP-1 promoter and suppress the function of hPR in human endometrial cells.

Previous studies have shown that progestin activates the transcription of IGFBP-1 (insulin-like growth factor binding protein-1). Four regions in the IGFBP-1 promotor have been identified to enhance the transcription. Two of the regions, located at -73 to -65 bp and -319 to -311 bp formed identical DNA-protein complexes with the nuclear extracts of endometrial stromal/decidual cells. To identify the binding protein(s) in endometrial cells that interact with these two regions, we have used the TGTCAATTA repeats (-319 to -11 bp of the IGFBP-1 promoter) to screen the human decidual cDNA library by yeast one-hybrid system. We found that Hox A10, HoxA11, HoxB2, HoxB4, and HoxD11 interacted with the TGTCAATTA repeats in yeast cells. Among these hox genes, the full-length coding region of HoxA10, HoxA11, and HoxB4 were used for functional analysis in three types of endometrial cells, undifferentiated endometrial stromal cells, decidual cells (differentiated stromal cells) and endometrial adenocarcinoma cell line (HEC1-B). All these endometrial cells produce IGFBP-1. Transient transfection assay showed that HoxA10 expression vector increased the promoter activity (the IGFBP-1 proximal promoter containing TGC/TCAATTA and two functional PRE sites) in endometrial stromal cells and in HEC-1B cells, but not in decidual cells. HoxB4 enhanced the promoter activity only in decidual cells, while HoxA11 had no apparent effect in all three types of cells. To evaluate whether Hox proteins would interact with progesterone receptor (hPR), cells were transfected with the promoter construct, Hox and hPR expression vectors. hPR alone activated the IGFBP-1 promoter activity, but expression of Hox gene suppressed the activation. Hox proteins also suppressed the hPR enhanced promoter activities of MMTV (containing consensus-PRE sites) and glycodelin (GdA, containing Sp1 site which mediates the hPR function). These data showed that Hox genes selectively activate the transcription of the IGFBP-1 and GdA genes in different types of endometrial cells. Hox genes, however, suppress the hPR enhanced activities. In addition, we found that HoxB4 expression was induced by estrogen and progestin. Other investigators have shown that HoxA10 and 11 were stimulated by progestin. These findings show that Hox proteins are molecular mediators of the steroid hormones during endometrial cell development.

Estrogen sulfotransferase in human placenta.

Human placental estrogen sulfotransferase (ESFT) was partially purified from the term placental cytosol by (NH4)2SO4 precipitation and agarose gel chromatography. Additional purifications caused a rapid loss of the enzyme activity. The activity was abolished by isoelectrofocusing but partially retained by chromatofocusing. The value of pI of human placental ESFT is 5.8 and the same value was obtained for bovine adrenal ESFT. The enzyme protein was able to bind to the affinity resin, estradiol-17-hemisuccinyl-1,2-diaminododecane sepharose 4B, but difficult to be extracted by estradiol (E2). The extract of the affinity resin showed one major protein band at 68,000 dalton on SDS-polyacrylamide gel electrophoresis. Kinetic studies using partially purified ESFT revealed that E2 is the best substrate for this enzyme. The relative rate of sulfurylation of E2, estrone, estriol and dehydroepiandrosterone at 4 microM (Km for E2) is 1, 0.3, 0.08 and 0.08, respectively.