- Email: [email protected]
Identification of the regulatory regions of the human aromatase P450 (CYP19) gene involved in placenta-specific expression
Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
Identification of the regulatory regions of the human aromatase P450 (CYP19) gene involved in placenta-specific expression夽 Amrita Kamat a , Carole R. Mendelson a,b,∗ b
a Department of Biochemistry, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Boulevard, Dallas, TX 75390, USA Department of Obstetrics–Gynecology, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Boulevard, Dallas, TX 75390, USA
Abstract Expression of the human CYP19 gene in placental syncytiotrophoblast, ovarian granulosa and luteal cells and adipose stromal cells is regulated by tissue-specific promoters which lie upstream of unique untranslated first exons. In placenta, the majority of CYP19 mRNA transcripts contain 5 -sequences encoded by exon I.1 which lies >35 kb upstream of the translation initiation sequence in exon II. Mononuclear cytotrophoblasts isolated from midterm human placenta spontaneously fuse in culture to form multinucleated syncytiotrophoblast. These morphological changes are associated with a marked induction of CYP19 gene expression. To functionally define genomic regions required for placenta-specific expression, fusion genes containing various amounts of exon I.1 5 -flanking sequence linked to the human growth hormone (hGH) structural gene, as reporter, were introduced into human trophoblast cells in primary monolayer culture and into transgenic mice. Our findings using transfected cells and transgenic mice suggest that sequences between −501 and −42 bp upstream of exon I.1 contain a positive enhancer element(s) that mediates the actions of trophoblast-specific transcription factors, as well as a negative element(s) that binds inhibitory transcription factors in other cell types. Our findings from transgenic studies further indicate that mouse placenta contains the necessary transcription factors required to activate the human CYP19 promoter although mouse placenta does not express endogenous aromatase. © 2002 Elsevier Science Ltd. All rights reserved. Keywords: Aromatase (CYP19); Trophoblast; Placenta; Gene expression
1. Introduction
1.1. Tissue-specific sites of expression
Aromatase, an enzyme complex of the endoplasmic reticulum, catalyzes the biosynthesis of C18 -estrogens (estradiol-17, estrone and estradiol) from C19 -steroids (testosterone, androstenedione and 16␣-hydroxyandrostenedione). This enzyme complex is comprised of the ubiquitous flavoprotein, NADPH-cytochrome P450 and a unique form of cytochrome P450 (P450arom, the product of the CYP19 gene) that is expressed exclusively in estrogen-producing cells [1–3]. In humans, there appears to be a single CYP19 gene [4]. Homozygous mutations of the CYP19 gene result in virilisation of the female fetus in utero and subsequent primary amenorrhea, whereas in males, there is continued linear bone growth beyond the time of puberty, delayed bone age and failure of epiphyseal closure [4].
In most vertebrates, including rodents, aromatase expression is restricted to the gonads and the brain [5]. However, in humans, aromatase is expressed in a number of tissues including the syncytiotrophoblast layer of the placenta [6], gonads [7,8], adipose tissue [9,10], brain [10,11], coronary arteries [12] and various fetal tissues including liver, skin and intestine [13–15]. In humans, ectopic expression of aromatase is known to occur in a number of pathophysiological states, including endometriosis, uterine leiomyomas, and in adipose tissues surrounding breast tumors [4]. During pregnancy, the human placenta provides the primary source of circulating estrogens after the 9th week of gestation by aromatisation of 16␣-hydroxylated C19 -steroids synthesized by the fetal adrenals [16]. At the present time, the physiological significance of the high levels of aromatase expression and estrogen biosynthesis in the human placenta are unclear. 1.2. Regulation of aromatase
夽 Proceedings of the Symposium: ‘Aromatase 2000 and the Third Generation’ (Port Douglas, Australia, 3–7 November 2000). ∗ Corresponding author. Tel.: +1-214-648-2944; fax: +1-214-648-8856. E-mail address: [email protected] (C.R. Mendelson).
CYP19 gene expression is subject to multifactorial regulation by a diverse group of hormones and factors including gonadotropins, cyclic AMP analogs, phorbol esters, growth
0960-0760/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 0 - 0 7 6 0 ( 0 1 ) 0 0 1 5 6 - X
174
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
factors and steroids that differ markedly in human ovarian, adipose, fetal liver and placental cells. Aromatase expression in the human ovary is regulated in granulosa and luteal cells at different phases of the ovarian cycle primarily by FSH and LH, respectively, which activate adenylyl cyclase and increase cyclic AMP formation [17]. In human adipose stromal cells in culture, aromatase activity and CYP19 gene expression are stimulated by glucocorticoids [9] acting in concert with cytokines [18]. In human fetal hepatocytes, aromatase activity also is stimulated by glucocorticoids in the presence of FCS [19]. By contrast, regulation of aromatase gene expression in human placenta is not very well understood. In primary cultures of placental cells, cyclic AMP analogues have been reported to have either no effect [20] or stimulatory effects [21] on aromatase activity. However, a number of stimulants including cyclic AMP have been reported to markedly induce aromatase activity in choriocarcinoma cell lines [22,23]. 1.3. CYP19 gene and its regulatory regions Human CYP19 is encoded by a single copy gene [4,24–26] that is localized on chromosome 15q21.1 [27]. The protein coding sequence is contained within nine exons (exons II–X) which span ∼35 kb of DNA [26]. The translation initiation and termination codons are present in exons II and X, respectively. The 5 -untranslated regions of CYP19 mRNAs in human ovary/testis, adipose stromal cells and placenta are encoded by alternative first exons which are located from ∼110 to ∼>35,000 bp upstream of the translation initiation site in exon II [5] (Fig. 1). Thus, CYP19 mRNAs in these tissues have distinct 5 -termini that are spliced onto a common site which lies ∼36 bp upstream of the translation initiation codon in exon II. It, therefore, is postulated that CYP19 gene expression in various tissues is driven by tissue-specific promoters that lie upstream of these exons. To our knowledge, CYP19 is the first member of the cytochrome P450 gene superfamily, and one of a limited number of eukaryotic genes, regulated by several different tissue-specific promoters. In the ovary, the 5 -untranslated region of CYP19 mRNA transcripts is encoded by an ovary-specific first exon that lies
immediately upstream of exon II while in human adipose tissue and adipose stromal cells, the CYP19 transcripts have 5 -ends encoded by exon I.4 which lies >20 kb upstream of exon II. However, alternative downstream first exons (I.3 and II) are used in human adipose stromal cells treated with cAMP and phorbol esters [28,29]. Interestingly, in fetal liver tissue, CYP19 transcripts containing exon I.4 predominate [13]. Since aromatase activity in cultured fetal hepatocytes also is stimulated by glucocorticoids in the presence of FCS [19], it is likely that promoter I.4 activity predominates in tissues in which CYP19 expression is regulated by glucocorticoids. On the other hand, in placenta, the majority of the CYP19 mRNA transcripts contain sequences encoded by exon I.1 [30,31] which lies >35 kb upstream of exon II. A minority of the transcripts in the placenta and fetal liver contain sequences encoded by exon I.2 which lies ∼9 kb upstream of exon II. Since the genomic clones containing exon I.1 and exon II do not overlap, the exact distance between them is not known [31]. The presence of the ovary-specific CYP19 promoter (promoter II) just upstream of the translation start site, in comparison to the distal major placental-specific promoter (promoter I.1) suggests that the latter was recruited much later in phylogeny with evolution of the primates. 1.4. Expression of CYP19 in human placenta Human placenta develops from the differentiating trophoectoderm cells at the periphery of the implanting blastocyst. As the blastocyst invades the endometrium, the trophoectoderm becomes polarized so that the chorion leave is formed by the outermost pole developing towards the endometrium. On the other hand, the innermost pole differentiates into the villous and extravillous parts of the placenta. The villous structures that develop are comprised of a core of mononuclear cytotrophoblasts, which are mitotically active. When these cells mature, they stop proliferating and fuse to form the syncytiotrophoblast layer that covers the placental villi and has numerous secretory and transport/processing functions [22]. Differentiation of cytotrophoblasts to syncytiotrophoblast generates a cascade of regulatory signals that result in the production of a diverse
Fig. 1. Structural organization of the human CYP19 gene and its alternative first exons. The protein coding sequences of the human CYP19 gene are shown as closed bars (exons II–X). The untranslated exons, I.1, I.4, I.2, indicated by open bars, are spliced onto a common acceptor site ∼36 bp upstream of the translation start site, ATG, in exon II. The heme binding region (HBR) is indicated in exon X, as are two alternative polyadenylation signals that give rise to two species of CYP19 transcripts of 3.4 and 2.9 kb. The genomic region shown spans a distance of 75 kb; however the true size is unknown [4].
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
array of polypeptide hormones, growth factors, steroids and steroid metabolizing enzymes, including aromatase. To elucidate the molecular events that promote and maintain syncytiotrophoblast differentiation, and culminate in expression of CYP19 gene in placenta, we have used both human placental cells in primary culture and transgenic technology. 1.4.1. Aromatase expression is induced in primary cultures of placental trophoblasts from midgestation human placenta in association with syncytiotrophoblast differentiation To define regions of the human CYP19 gene that mediate tissue-specific expression in placenta, we have established a primary cell culture system using purified cytotrophoblast cells from midtrimester human placenta [36,37]. Freshly isolated cytotrophoblast cells were found to have very low or undetectable levels of aromatase activity (Fig. 2A) and CYP19 mRNA levels (Fig. 2B) prior to culture. However, after culture, these cells differentiate spontaneously to form multinuclear syncytiotrophoblast [37,38]. Aromatase activity and CYP19 mRNA levels are detectable within 24 h after plating and continue to increase further as a function of time in culture (Fig. 2). These findings, therefore, indicate that differentiation of placental cytotrophoblasts to syncytiotrophoblast is associated with a marked induction of aromatase activity and CYP19 gene expression [36]. 1.4.2. Aromatase expression in choriocarcinoma cell lines As mentioned earlier, several laboratories have used choriocarcinoma cell lines like JEG-3, JAR and BeWo cells for
175
studying placental endocrine function and gene expression [23,24,32–35,39,40]. Like normal trophoblast cells, BeWo, JAR and JEG-3 cell lines lack 17␣-hydroxylase/17,20 lyase activity but produce human chorionic gonadotropins, progesterone and aromatize exogenous androgens into estrogens. However, unlike normal trophoblast cells, these cell lines show only modest degrees of differentiation and secrete low amounts of hormones characteristic of syncytiotrophoblast, such as chorionic somatomammotropin [22,39,40]. Studies by Taylor et al. [41] have shown that BeWo cells have cytotrophoblast-like features and express low endogenous levels of aromatase. When cultured in presence of forskolin, a direct activator of adenylyl cyclase, these cells acquire an intermediate trophoblast phenotype with a ∼13.5-fold increase in aromatase activity. On the other hand, JEG-3 cells do not show any cytological changes in the presence of cAMP although there is an increase in aromatase activity [22]. We performed our initial studies in JEG-3 cells because of their ease of transfection, availability and rapid proliferation. However, we found that the JEG-3 cells exhibit aromatase activity that is approximately 1/50th that of primary cultures of human placental cells; CYP19 mRNA levels are undetectable by Northern blotting (Fig. 3) [36]. In consideration of this finding, we believe that the human trophoblast cells in primary culture provide a more physiologically relevant model system for mapping genomic regions required for placenta-specific expression of the human CYP19 gene than transfected human choriocarcinoma cells which express very low levels of aromatase.
Fig. 2. Aromatase activity (A) and Northern blot analysis of CYP19 mRNA in human trophoblasts before and after differentiation to syncytiotrophoblast in culture. (A) Freshly isolated cytotrophoblasts were suspended in DMEM containing 2% fetal bovine serum and plated at a density of 2 × 106 cells per dish. The aromatase activity (pmol androgen metabolized to estrogen/mg protein/min) was assayed over a 4-day-period by the incorporation of tritium from [1-3 H]androstenedione into water. The values are the mean ± S.E.M. of data from triplicate dishes of cells. (B) Total RNA (20 g per lane) obtained from freshly isolated cytotrophoblasts and syncytiotrophoblast after 24, 48 and 72 h in culture were analyzed for CYP19 mRNA by Northern blotting using a full-length 32 P-labeled human CYP19 cDNA probe [36].
176
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
Fig. 3. Aromatase activity (A) and Northern blot analysis of CYP19 mRNA (B) in human trophoblasts and JEG-3 cells. (A) Aromatase activity (pmol androgen metabolized to estrogen/mg protein/min) in trophoblast cells in primary culture and in JEG-3 cells was assayed by analysis of the incorporation of tritium from [1-3 H]androstenedione into water. Values shown are the mean ± S.E.M. of aromatase activity of triplicate dishes of cells on day 4 of culture. (B) Total RNA (20 g) obtained from freshly isolated cytotrophoblasts, syncytiotrophoblast after 4 days in culture and JEG-3 cells at 80–90% confluence was analyzed for CYP19 mRNA by Northern blotting using a full-length 32 P-labeled human CYP19 cDNA as probe [36].
1.5. Analysis of genomic sequences that mediate placenta-specific expression of CYP19 using cultured placental cells 1.5.1. Analysis of genomic sequences that mediate placenta-specific expression of CYP19 using primary cultures of human trophoblast cells To functionally map genomic sequences required for placenta-specific CYP19 gene expression, we have chosen to use primary cultures of human trophoblast cells which express high levels of aromatase. Fusion genes comprised of various amounts of CYP19 exon I.1 5 -flanking DNA, linked to the human growth hormone (hGH) structural gene, as reporter, were incorporated into a replication-defective human adenovirus vector (Ad5) and introduced into primary cultures of cytotrophoblasts and other cell types by infection. This results in highly efficient and reproducible transfer of fusion gene constructs into cells [36,42,43]. We observed that expression of CYP19I.1−2400 , CYP19I.1−923 and CYP19I.1−501 :hGH fusion genes increased in primary human trophoblasts in culture in concert with syncytiotrophoblast differentiation and induction of CYP19 gene expression. Expression of these fusion genes was undetectable in non-placental cells. On the other hand, we found that fusion genes containing 246, 201 and 125 bp of exon I.1 5 -flanking DNA were expressed in primary trophoblast cultures as well as in non-placental cells. By contrast, expression of the 42 bp containing fusion gene construct which includes only the minimal promoter region, remained essentially undetectable in placental and non-placental cells (Fig. 4). Taken together, all these results suggest that the region between −501 and −42 bp upstream from exon I.1 is sufficient for enhanced levels of CYP19 promoter I.1 activity in primary trophoblast cells in culture
and that the sequences between −501 and −246 bp contain silencer elements that may bind inhibitory transcription factors in non-placental cell types [36]. Further studies using deletion analysis, site-directed mutagenesis and electrophoretic mobility shift assays have indicated that two overlapping hexameric sequences (AGGTCA, −183 and −191 bp) which may bind a member(s) of the nuclear hormone receptor superfamily [35] and a G/C-rich sequence (−233 bp) which binds Sp1 [44] and other, as yet, unidentified transcription factors, are critical for maximum induction of CYP19I.1 promoter activity during syncytiotrophoblast differentiation [36]. 1.5.2. Analysis of genomic sequences that mediate placenta-specific expression of CYP19 using choriocarcinoma cell lines Toda et al. [32] have characterized two cis-acting elements by transient transfection of human BeWo choriocarcinoma cells, termed as hATRE-1 (human aromatase cytochrome P450 gene transcriptional regulatory element-1) and hATRE-2 which lie between −2238 and −2214 bp, and between −2141 and −2115 bp upstream of the transcriptional start site in exon I.1, respectively. Further characterization of these sites indicated that hATRE-1 acts as a repressor whereas the region between −2141 and −2115 bp acts as an enhancer, in response to 12-O-tetradecanoylphorbol-13-acetate and binds the transcription factor, C/EBP. Additionally, they have localized a cell-type specific enhancer element between −242 and −166 bp relative to the transcriptional start site in exon I.1 that is necessary for maximum transcriptional enhancement of aromatase in BeWo cells. Using JEG-3 choriocarcinoma cells to analyze the 5 -flanking region of CYP19 gene exon I.1, Yamada et al. [23] have reported that the initial 301 bp upstream of exon
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
177
Fig. 4. Expression of CYP19I.1:hGH fusion genes in primary human trophoblast cultures and in various cell lines. The different cell types shown were infected with 1 × 106 recombinant infectious viral particles containing CYP19I.1−2400 :hGH, CYP19I.1−923 :hGH, CYP19I.1−501 :hGH, CYP19I.1−246 :hGH, CYP19I.1−201 :hGH, CYP19I.1−125 :hGH and CYP19I.1−42 :fusion genes. Shown are the levels of hGH that accumulated in the medium over a 24-h-period between days 2 and 3 of culture. Values are the mean ± S.E.M. of data from 2–4 independent experiments, each conducted in triplicate [36].
I.1 may confer placenta-specific expression to the CYP19 gene. Recently, they have cloned a glial cells missing (GCM) motif protein that binds to a trophoblast-specific element within this region and may be involved in the regulation of CYP19 and possibly various other placenta-specific genes [34]. However, we have observed that the CYP19I.1:hGH fusion genes express at much lower levels in the JEG-3 cells than the primary cultures. Additionally, their pattern of expression is not similar to that seen in the primary trophoblast cells in culture [36]. Hence, although JEG-3 cells express endogenous aromatase and share many other characteristics of normal trophoblast cells, they may not provide a relevant system for delineating genomic sequences required for placenta-specific expression of the human CYP19 gene. 1.6. Use of transgenic mice to analyze regions of CYP19 gene required for placenta-specific expression Transgenic mice provide the most relevant model system to map genomic regions required for appropriate tissue-specific and developmental regulation of gene expression. Initially, we were uncertain whether the mouse would serve as an appropriate model system, since rodent placenta does not express aromatase. However, we observed that expression of CYP19I.1−2400 :hGH or CYP19I.1−501 :hGH fusion genes in transgenic mice was placenta-specific and developmentally regulated. These results clearly indicate that mouse placenta contains the necessary transcription factors to activate the human CYP19 gene placental-specific promoter but lacks the critical cis-acting elements required
for expression of the endogenous CYP19 gene. Similar studies on the ␣-subunit of the glycoprotein hormones, which is expressed in the placentae of only primates and horses [46,47], have indicated that lack of expression of the glycoprotein hormone ␣-subunit in rodent placenta was due to a single nucleotide change in a cyclic AMP-response element (CRE) and not due to the absence of essential regulatory transcription factors. However, unlike the ␣-subunit of the glycoprotein hormones, which appears to use a single promoter, the complexity of the CYP19 gene has made it difficult to define the genetic basis for placenta-specific expression of human aromatase by comparison of promoter sequences. Humans, cattle, horses and pigs, all appear to utilize placenta-specific promoters for tissue-specific expression of the CYP19 gene. However, in those species in which the 5 -flanking regions of the gene have been cloned, there is very low sequence identity and the placenta-specific promoters lie at variable distances upstream of the translation start site in exon II [48,49]. To localize CYP19I.1−501 :hGH fusion gene expression within the mouse placenta, in situ hybridization was performed using an antisense hGH probe. Reporter gene expression was absent in the spongiotrophoblast and the trophoblast giant cells, which produce many of the placental polypeptide hormones and steroidogenic enzymes [50,51]. On the other hand, the transgene was expressed as early as E10.5 specifically in the mouse labyrinthine trophoblast layer (Fig. 5.) [45]. This layer which is highly vascularized is trilaminar; its outer cells, which cover two layers of
178
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
Fig. 5. Expression of a CYP19I.1−501 :hGH fusion gene is restricted to labyrinthine trophoblast in transgenic mice. Placental tissues obtained from E10.5 or E17.5 fetal transgenic mice carrying the CYP19I.1−501 :hGH fusion gene (A–C) were processed for in situ hybridization by using a 35 S-labeled antisense hGH cRNA probe and exposed to photographic emulsion for 1–2 weeks. Bright and dark field microscopy then were performed. (A) Dark field micrograph of a placental tissue section from an E10.5 transgenic mouse carrying the CYP19I.1−501 :hGH fusion gene hybridized with radiolabeled hGH antisense cRNA probe. Silver grains indicating hGH mRNA transcripts appear as white spots. (B) Bright field micrograph of the haematoxylin-stained E10.5 placental tissue section shown in A. (C) Dark field micrograph of E17.5 placental tissue section incubated with radiolabeled hGH antisense cRNA probe. (D) Dark field micrograph of E17.5 placental section from a nontransgenic mouse incubated with radiolabeled hGH antisense cRNA probe (gc: trophoblast giant cell; sp: spongiotrophoblast; lab: labyrinthine trophoblast; ud: uterine decidua) [45].
syncytial cells, are bathed in maternal blood and thus play an important role in nutrient and gas exchange. In this respect, the labyrinthine layer of the mouse placenta, is most analogous to the human syncytiotrophoblast, which expresses aromatase and also is bathed in maternal blood [52]. These interesting findings suggest that 501 bp of exon I.1 5 -flanking sequence is sufficient for labyrinthine-specific expression of the human aromatase promoter in mouse.
2. Conclusions and future perspectives To functionally define genomic regions that direct appropriate cell-/tissue-specific and developmental expression of human CYP19 gene in placenta, studies in transgenic mice and primary cultures of human trophoblast cells were undertaken. Studies using primary cultures of human placental cells transfected with the CYP19I.1:hGH fusion genes suggest that the region between 501 and 42 bp upstream of exon I.1 contains response elements required for trophoblast-specific expression of the CYP19 gene, whereas
the region between −501 and −246 bp of exon I.1 may bind transcriptional repressors that prevent CYP19 gene expression in non-placental cells. In transgenic mice, a relatively small region (∼500 bp) >40,000 bp upstream of the protein coding region of the human CYP19 gene is able to direct expression in placenta in an appropriate cell-specific manner. Although, mouse placenta does not express aromatase, these findings indicate that transgenic mice can serve as an appropriate model system for defining regulatory regions surrounding the human CYP19 gene that mediate trophoblast-specific expression. These results also suggest that the transcription factors that direct placenta-specific expression of CYP19 are highly conserved among the animal species, although the response elements that bind these factors probably are not. Furthermore, the finding that transgene expression is confined to the labyrinthine trophoblast suggests that this region of the mouse placenta is most analogous to the human syncytiotrophoblast, which expresses high levels of aromatase. Studies are currently in progress to identify the critical cis-acting elements and characterize the transcription factors
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
that bind to these sites and direct appropriate expression of CYP19. This research, therefore, will provide insight into the molecular events that promote and maintain syncytiotrophoblast differentiation and culminate in placenta-specific expression of CYP19.
Acknowledgements This research was supported by NIH Grant 5 RO1 DK-31206. A.K. was supported in part by NIH Training Grant 5-T32-HD-07190-16.
References [1] C.R. Mendelson, E.E. Wright, J.C. Porter, C.T. Evans, E.R. Simpson, Preparation and characterization of polyclonal and monoclonal antibodies against human aromatase cytochrome P450 (P450arom) and their use in its purification, Arch. Biochem. Biophys. 243 (1985) 480–491. [2] S. Nakajin, M. Shimoda, P.F. Hall, Purification to homogeneity of aromatase from human placenta, Biochem. Biophys. Res. Commun. 134 (1986) 704–710. [3] J.T. Kellis Jr., L.E. Vickery, Purification and characterization of human placental aromatase cytochrome P450, J. Biol. Chem. 262 (1987) 4413–4420. [4] E.R. Simpson, Y. Zhao, V.R. Agarwal, M.D. Michael, S.E. Bulun, M.M. Hinshelwood, S. Graham-Lorence, T. Sun, C.R. Fisher, K. Qin, C.R. Mendelson, Aromatase expression in health and disease, Rec. Prog. Hormone Res. 52 (1997) 185–214. [5] E.R. Simpson, M.S. Mahendroo, G.D. Means, M.W. Kilgore, M.M. Hinshelwood, S. Graham-Lorence, B. Amarneh, Y. Ito, C.R. Fisher, M.D. Michael, et al., Aromatase cytochrome P450, the enzyme responsible for estrogen biosynthesis, Endocrine Rev. 15 (3) (1994) 342–355. [6] N. Fournet-Dulguerov, N.J. MacLusky, C.Z. Leranth, R. Todd, C.R. Mendelson, E.R. Simpson, F. F Naftolin, Immunohistochemical localization of aromatase cytochrome P450 and estradiol dehydrogenase in the syncytiotrophoblast of the human placenta, J. Clin. Endocrinol. Metabol. 65 (1987) 757–764. [7] K.P. McNatty, D.T. Baird, A. Bolton, P. Chambers, C.S. Corker, H. McLean, Concentration of oestrogens and androgens in human ovarian venous plasma and follicular fluid throughout the menstrual cycle, J. Endocrinol. 71 (1) (1976) 77–85. [8] S. Carreau, B. Biluska, J. Levallet, Male germ cells: a new source of estrogens in the mammalian testis, Annales d’Endocrinologie 59 (2) (1998) 79–92. [9] E.R. Simpson, G.E. Ackerman, M.E. Smith, C.R. Mendelson, Estrogen formation in stromal cells of adipose tissue of women: induction by glucocorticoids, Proc. Natl. Acad. Sci. U.S.A. 78 (1981) 5690–5694. [10] J.M. Grodin, P.K. Siiteri, P.C. MacDonald, Source of estrogen production in postmenopausal women, J. Clin. Endocrinol. Metabol. 36 (1973) 207–214. [11] F. Naftolin, K.J. Ryan, I.J. Davies, V.V. Reddy, F. Flores, Z. Petro, M. Kuhn, R.J. White, Y. Takaoka, L. Wolin, The formation of estrogens by central neuroendocrine tissues, Recent Prog. Hormone Res. 31 (1975) 295–319. [12] S. Diano, T.L. Horvath, G. Mor, T. Register, M. Adams, N. Harada, F. Naftolin, Aromatase and estrogen receptor immunoreactivity in the coronory arteries of monkeys and human subjects, Menopause 6 (1) (1999) 21–28.
179
[13] K. Toda, E.R. Simpson, C.R. Mendelson, Y. Shizuta, M.W. Kilgore, Expression of the gene encoding aromatase cytochrome P450 (CYP19) in fetal tissues, Mol. Endocrinol. 8 (1994) 210–217. [14] K.J. Doody, B.R. Carr, Aromatase in human fetal tissues, Am. J. Obstet. Gynecol. 10 (1984) 241–244. [15] T. Price, J. Aitkin, E.R. Simpson, Relative expression of aromatase cytochrome P450 in human fetal tissues as determined by competitive polymerase chain reaction (PCR) amplification, J. Clin. Endocrinol. Metabol. 74 (1992) 879–883. [16] R.B. Everett, P.C. MacDonald, Endocrinology of the placenta, Ann. Rev. Med. 30 (1979) 473–488. [17] M.P. Steinkampf, C.R. Mendelson, E.R. Simpson, Regulation by FSH of the synthesis of aromatase cytochrome P-450 (P450arom) in human granulosa cells, Mol. Endo. 1 (1987) 465–471. [18] Y. Zhao, J.E. Nichols, S.E. Bulun, C.R. Mendelson, E.R. Simpson, Aromatase P450 gene expression in human adipose tissue. Role of a Jak/STAT pathway in regulation of adipose-specific promoter activity, J. Biol. Chem. 270 (1995) 16449–16457. [19] M.J. Lanoux, W.H. Cleland, C.R. Mendelson, B.R. Carr, E.R. Simpson, Factors affecting the conversion of androstenedione to estrogens by human fetal hepatocytes in primary monolayer culture, Endocrinology 117 (1985) 361–368. [20] J.E. Nestler, Modulation of aromatase and P450 cholesterol side-chain cleavage enzyme activities in human placental cytotrophoblasts by insulin and insulin-like growth factor I, Endocrinology 121 (5) (1987) 1845–1852. [21] J. Lobo, F.L. Bellino, Estrogen synthetase (aromatase) activity in primary term placental cells: effects of cell preparation, growth medium and serum on adenosine 3 , 5 -monophosphate response, J. Clin. Enocrinol. Metabol. 69 (1989) 868–874. [22] G.E. Ringler, J.F. Strauss, In vitro systems for the study of human placental endocrine function, Endocrine Rev. 11 (1990) 105–123. [23] K. Yamada, N. Harada, S. Honda, Y. Takagi, Regulation of placenta-specific expression of the aromatase cytochrome P450 gene, J. Biol. Chem. 270 (1995) 25064–25069. [24] K. Toda, M. Terashima, T. Kawamoto, H. Sumimoto, Y. Yokoyama, I. Kuribayashi, Y. Mitsuuchi, T. Maeda, Y. Yamamoto, Y. Sagara, H. Ikeda, Y. Shizuta, Structural and functional characterization of human aromatase P450 gene, Eur. J. Biochem. 193 (1990) 559–565. [25] N. Harada, K. Yamada, K. Saito, N. Kibe, S. Dohmae, Y. Takagi, Structural characterization of the human estrogen synthetase (aromatase) gene, Biochem. Biophys. Res. Commun. 166 (1990) 365–372. [26] G.D. Means, M.S. Mahendroo, J.C. Corbin, J.M. Mathis, F.E. Powell, C.R. Mendelson, E.R. Simpson, Structural analysis of the gene encoding human aromatase cytochrome P-450, the enzyme responsible for estrogen biosynthesis, J. Biol. Chem. 264 (1989) 19385–19391. [27] S. Chen, M.J. Besman, R.S. Sparkes, S. Zollman, I. Klisak, T. Mohandes, P.F. Hall, J.E. Shively, Human aromatase: cDNA cloning, southern blot analysis and assignment of the gene to chromosome 15, DNA 7 (1988) 27–38. [28] M.S. Mahendroo, C.R. Mendelson, E.R. Simpson, Tissue-specific and hormonally-controlled alternative promoters regulate aromatase cytochrome P450 gene expression in human adipose tissue, J. Biol. Chem. 268 (1993) 19463–19470. [29] Y. Zhao, C.R. Mendelson, E.R. Simpson, Characterization of the sequences of the human CYP19 (aromatase) gene that mediate regulation by glucocorticoids in adipose stromal cells and fetal hepatocytes, Mol. Endocrinol. 9 (1995) 340–349. [30] M.S. Mahendroo, G.D. Means, C.R. Mendelson, E.R. Simpson, Tissue-specific expression of human P450arom: the promoter responsible for expression in adipose is different from that utilized in placenta, J. Biol. Chem. 266 (1991) 11276–11281. [31] G.D. Means, M.W. Kilgore, M.S. Mahendroo, C.R. Mendelson, E.R. Simpson, Tissue-specific promoters regulate aromatase cytochrome P450 gene expression in human ovary and fetal tissues, Mol. Endocrinol. 5 (1991) 2005–2013.
180
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
[32] K. Toda, Y. Shizuta, Identification and characterization of cis-acting regulatory elements for the expression of the human aromatase cytochrome P-450 gene, J. Biol. Chem. 269 (1994) 8099–8107. [33] K. Toda, L.X. Yang, Y. Shizuta, Transcriptional regulation of the human aromatase cytochrome P450 gene expression in human placental cells, J. Steroid. Biochem. Mol. Biol. 53 (1995) 181–190. [34] K. Yamada, S. Honda, N. Harada, T. Okazaki, A GCM motif protein is involved in placenta-specific expression of human aromatase gene, J. Biol. Chem. 274 (1999) 32279–32286. [35] T. Sun, Y. Zhao, D.J. Mangelsdorf, E.R. Simpson, Characterization of a region upstream of exon I.1 of the human CYP19 (aromatase) gene that mediates regulation by retinoids in human choriocarcinoma cells, Endocrinology 139 (1998) 1684–1691. [36] A. Kamat, J.L. Alcorn, C. Kunczt, C.R. Mendelson, Characterization of the regulatory regions of the human aromatase (P450arom) gene involved in placenta-specific expression, Mol. Endocrinol. 12 (1998) 1764–1777. [37] H.J. Kliman, J.E. Nestler, E. Sermasi, J.M. Sanger, J.F. Strauss, Purification, characterization and in vitro differentiation of cytotrophoblasts from human term placentae, Endocrinology 118 (1986) 1567–1582. [38] J.O. Lobo, F.L. Bellino, L. Bankert, Estrogen synthetase activity in human term placental cells in monolayer culture, Endocrinology 116 (1985) 889–895. [39] H.-P. Hohn, M. Linke, B. Ugele, H.-W. Denker, Differentiation markers and invasiveness: discordent regulation in normal trophoblast and choriocarcinoma cells, Exp. Cell Res. 244 (1998) 249–258. [40] R.D. Bahn, A. Worsham, K.V. Speeg, M. Ascoli, D. Rabin, Characterization of steroid production in cultured human choriocarcinoma cells, J. Clin. Endocrinol. Metabol. 52 (1981) 447–450. [41] R.N. Taylor, E.D. Newman, S. Chen, Forskolin and methotrexate induce an intermediate trophoblast phenotype in cultured human choriocarcinoma cells, Am. J. Obstet. Gynecol. 164 (1991) 204–210. [42] B.L. Davidson, E.D. Allen, K.F. Kozarsky, J.M. Wilson, J. Roessler, A model system for in vivo gene transfer into the central nervous system using an adenoviral vector, Nat. Genet. 3 (1993) 219–223. [43] J.L. Alcorn, E. Gao, Q. Chen, M.E. Smith, R.D. Gerard, C.R. Mendelson, Genomic elements involved in transcriptional regulation
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
of the rabbit surfactant protein-A gene, Mol. Endocrinol. 7 (1993) 1072–1085. M.R. Briggs, J.T. Kadonaga, S.P. Bell, R. Tijan, Purification and biochemical characterization of the promoter-specific transcription factor Sp1, Science 234 (1986) 47–52. A. Kamat, K.H. Graves, M.E. Smith, J.A. Richardson, C.R. Mendelson, A 500 bp region, ∼40 kb upstream of the human CYP19 (aromatase) gene, mediates placenta-specific expression in transgenic mice, Proc. Natl. Acad. Sci. U.S.A. 96 (1999) 4575–4580. J.A. Bokar, R.A. Keri, T.A. Farmerie, R.A. Fenstermaker, B.A. Andersen, D.L. Hamernik, J. Yun, T. Wagner, J.H. Nilson, Expression of the glycoprotein hormone ␣-subunit gene in the placenta requires a functional cyclic AMP response element, whereas a different cis-acting element mediates pituitary-specific expression, Mol. Cell Biol. 9 (11) (1989) 5113–5122. J.H. Nilson, J.A. Bokar, C.M. Clay, T.A. Farmerie, R.A. Fenstermaker, D.L. Hamernik, R.A. Keri, Different combinations of regulatory elements may explain why placenta-specific expression of the glycoprotein hormone ␣-subunit gene occurs only in primates and horses, Biol. Reprod. 44 (1991) 231–237. M.M. Hinshelwood, Z. Liu, A.J. Conley, E.R. Simpson, Demonstration of tissue-specific promoters in non-primate species that express aromatase P450 in placentae, Biol. Reprod. 53 (5) (1995) 1151–1159. I. Choi, D.L. Troyer, D.L. Cornwell, K.R. Kirby-Dobbels, W.R. Collante, F.A. Simmen, Closely related genes encode developmental and tissue isoforms of porcine cytochrome P450 aromatase, DNA Cell Biol. 16 (1997) 769–777. M.J. Soares, B.M. Chapman, C.A. Rasmussen, G. Dai, T. Kamei, K.E. Orwig, Differentiation of trophoblast endocrine cells, Placenta 17 (1996) 277–289. S.T. Suhr, J.O. Rahal, K.E. Mayo, Mouse growth-hormone releasing hormone: precursor structure and expression in brain and placenta, Mol. Endocrinol. 3 (1989) 1693–1700. W.D. Billington, Organization, ultrastructure and histochemistry of the placenta: immunological considerations, in: R.G. Edwards, C.W.S. Howe, M.H. Johnson (Eds.), Immunobiology of Trophoblast, Cambridge University Press, Cambridge, 1975, pp. 67–85.
Identification of the regulatory regions of the human aromatase P450 (CYP19) gene involved in placenta-specific expression夽 Amrita Kamat a , Carole R. Mendelson a,b,∗ b
a Department of Biochemistry, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Boulevard, Dallas, TX 75390, USA Department of Obstetrics–Gynecology, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Boulevard, Dallas, TX 75390, USA
Abstract Expression of the human CYP19 gene in placental syncytiotrophoblast, ovarian granulosa and luteal cells and adipose stromal cells is regulated by tissue-specific promoters which lie upstream of unique untranslated first exons. In placenta, the majority of CYP19 mRNA transcripts contain 5 -sequences encoded by exon I.1 which lies >35 kb upstream of the translation initiation sequence in exon II. Mononuclear cytotrophoblasts isolated from midterm human placenta spontaneously fuse in culture to form multinucleated syncytiotrophoblast. These morphological changes are associated with a marked induction of CYP19 gene expression. To functionally define genomic regions required for placenta-specific expression, fusion genes containing various amounts of exon I.1 5 -flanking sequence linked to the human growth hormone (hGH) structural gene, as reporter, were introduced into human trophoblast cells in primary monolayer culture and into transgenic mice. Our findings using transfected cells and transgenic mice suggest that sequences between −501 and −42 bp upstream of exon I.1 contain a positive enhancer element(s) that mediates the actions of trophoblast-specific transcription factors, as well as a negative element(s) that binds inhibitory transcription factors in other cell types. Our findings from transgenic studies further indicate that mouse placenta contains the necessary transcription factors required to activate the human CYP19 promoter although mouse placenta does not express endogenous aromatase. © 2002 Elsevier Science Ltd. All rights reserved. Keywords: Aromatase (CYP19); Trophoblast; Placenta; Gene expression
1. Introduction
1.1. Tissue-specific sites of expression
Aromatase, an enzyme complex of the endoplasmic reticulum, catalyzes the biosynthesis of C18 -estrogens (estradiol-17, estrone and estradiol) from C19 -steroids (testosterone, androstenedione and 16␣-hydroxyandrostenedione). This enzyme complex is comprised of the ubiquitous flavoprotein, NADPH-cytochrome P450 and a unique form of cytochrome P450 (P450arom, the product of the CYP19 gene) that is expressed exclusively in estrogen-producing cells [1–3]. In humans, there appears to be a single CYP19 gene [4]. Homozygous mutations of the CYP19 gene result in virilisation of the female fetus in utero and subsequent primary amenorrhea, whereas in males, there is continued linear bone growth beyond the time of puberty, delayed bone age and failure of epiphyseal closure [4].
In most vertebrates, including rodents, aromatase expression is restricted to the gonads and the brain [5]. However, in humans, aromatase is expressed in a number of tissues including the syncytiotrophoblast layer of the placenta [6], gonads [7,8], adipose tissue [9,10], brain [10,11], coronary arteries [12] and various fetal tissues including liver, skin and intestine [13–15]. In humans, ectopic expression of aromatase is known to occur in a number of pathophysiological states, including endometriosis, uterine leiomyomas, and in adipose tissues surrounding breast tumors [4]. During pregnancy, the human placenta provides the primary source of circulating estrogens after the 9th week of gestation by aromatisation of 16␣-hydroxylated C19 -steroids synthesized by the fetal adrenals [16]. At the present time, the physiological significance of the high levels of aromatase expression and estrogen biosynthesis in the human placenta are unclear. 1.2. Regulation of aromatase
夽 Proceedings of the Symposium: ‘Aromatase 2000 and the Third Generation’ (Port Douglas, Australia, 3–7 November 2000). ∗ Corresponding author. Tel.: +1-214-648-2944; fax: +1-214-648-8856. E-mail address: [email protected] (C.R. Mendelson).
CYP19 gene expression is subject to multifactorial regulation by a diverse group of hormones and factors including gonadotropins, cyclic AMP analogs, phorbol esters, growth
0960-0760/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 0 - 0 7 6 0 ( 0 1 ) 0 0 1 5 6 - X
174
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
factors and steroids that differ markedly in human ovarian, adipose, fetal liver and placental cells. Aromatase expression in the human ovary is regulated in granulosa and luteal cells at different phases of the ovarian cycle primarily by FSH and LH, respectively, which activate adenylyl cyclase and increase cyclic AMP formation [17]. In human adipose stromal cells in culture, aromatase activity and CYP19 gene expression are stimulated by glucocorticoids [9] acting in concert with cytokines [18]. In human fetal hepatocytes, aromatase activity also is stimulated by glucocorticoids in the presence of FCS [19]. By contrast, regulation of aromatase gene expression in human placenta is not very well understood. In primary cultures of placental cells, cyclic AMP analogues have been reported to have either no effect [20] or stimulatory effects [21] on aromatase activity. However, a number of stimulants including cyclic AMP have been reported to markedly induce aromatase activity in choriocarcinoma cell lines [22,23]. 1.3. CYP19 gene and its regulatory regions Human CYP19 is encoded by a single copy gene [4,24–26] that is localized on chromosome 15q21.1 [27]. The protein coding sequence is contained within nine exons (exons II–X) which span ∼35 kb of DNA [26]. The translation initiation and termination codons are present in exons II and X, respectively. The 5 -untranslated regions of CYP19 mRNAs in human ovary/testis, adipose stromal cells and placenta are encoded by alternative first exons which are located from ∼110 to ∼>35,000 bp upstream of the translation initiation site in exon II [5] (Fig. 1). Thus, CYP19 mRNAs in these tissues have distinct 5 -termini that are spliced onto a common site which lies ∼36 bp upstream of the translation initiation codon in exon II. It, therefore, is postulated that CYP19 gene expression in various tissues is driven by tissue-specific promoters that lie upstream of these exons. To our knowledge, CYP19 is the first member of the cytochrome P450 gene superfamily, and one of a limited number of eukaryotic genes, regulated by several different tissue-specific promoters. In the ovary, the 5 -untranslated region of CYP19 mRNA transcripts is encoded by an ovary-specific first exon that lies
immediately upstream of exon II while in human adipose tissue and adipose stromal cells, the CYP19 transcripts have 5 -ends encoded by exon I.4 which lies >20 kb upstream of exon II. However, alternative downstream first exons (I.3 and II) are used in human adipose stromal cells treated with cAMP and phorbol esters [28,29]. Interestingly, in fetal liver tissue, CYP19 transcripts containing exon I.4 predominate [13]. Since aromatase activity in cultured fetal hepatocytes also is stimulated by glucocorticoids in the presence of FCS [19], it is likely that promoter I.4 activity predominates in tissues in which CYP19 expression is regulated by glucocorticoids. On the other hand, in placenta, the majority of the CYP19 mRNA transcripts contain sequences encoded by exon I.1 [30,31] which lies >35 kb upstream of exon II. A minority of the transcripts in the placenta and fetal liver contain sequences encoded by exon I.2 which lies ∼9 kb upstream of exon II. Since the genomic clones containing exon I.1 and exon II do not overlap, the exact distance between them is not known [31]. The presence of the ovary-specific CYP19 promoter (promoter II) just upstream of the translation start site, in comparison to the distal major placental-specific promoter (promoter I.1) suggests that the latter was recruited much later in phylogeny with evolution of the primates. 1.4. Expression of CYP19 in human placenta Human placenta develops from the differentiating trophoectoderm cells at the periphery of the implanting blastocyst. As the blastocyst invades the endometrium, the trophoectoderm becomes polarized so that the chorion leave is formed by the outermost pole developing towards the endometrium. On the other hand, the innermost pole differentiates into the villous and extravillous parts of the placenta. The villous structures that develop are comprised of a core of mononuclear cytotrophoblasts, which are mitotically active. When these cells mature, they stop proliferating and fuse to form the syncytiotrophoblast layer that covers the placental villi and has numerous secretory and transport/processing functions [22]. Differentiation of cytotrophoblasts to syncytiotrophoblast generates a cascade of regulatory signals that result in the production of a diverse
Fig. 1. Structural organization of the human CYP19 gene and its alternative first exons. The protein coding sequences of the human CYP19 gene are shown as closed bars (exons II–X). The untranslated exons, I.1, I.4, I.2, indicated by open bars, are spliced onto a common acceptor site ∼36 bp upstream of the translation start site, ATG, in exon II. The heme binding region (HBR) is indicated in exon X, as are two alternative polyadenylation signals that give rise to two species of CYP19 transcripts of 3.4 and 2.9 kb. The genomic region shown spans a distance of 75 kb; however the true size is unknown [4].
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
array of polypeptide hormones, growth factors, steroids and steroid metabolizing enzymes, including aromatase. To elucidate the molecular events that promote and maintain syncytiotrophoblast differentiation, and culminate in expression of CYP19 gene in placenta, we have used both human placental cells in primary culture and transgenic technology. 1.4.1. Aromatase expression is induced in primary cultures of placental trophoblasts from midgestation human placenta in association with syncytiotrophoblast differentiation To define regions of the human CYP19 gene that mediate tissue-specific expression in placenta, we have established a primary cell culture system using purified cytotrophoblast cells from midtrimester human placenta [36,37]. Freshly isolated cytotrophoblast cells were found to have very low or undetectable levels of aromatase activity (Fig. 2A) and CYP19 mRNA levels (Fig. 2B) prior to culture. However, after culture, these cells differentiate spontaneously to form multinuclear syncytiotrophoblast [37,38]. Aromatase activity and CYP19 mRNA levels are detectable within 24 h after plating and continue to increase further as a function of time in culture (Fig. 2). These findings, therefore, indicate that differentiation of placental cytotrophoblasts to syncytiotrophoblast is associated with a marked induction of aromatase activity and CYP19 gene expression [36]. 1.4.2. Aromatase expression in choriocarcinoma cell lines As mentioned earlier, several laboratories have used choriocarcinoma cell lines like JEG-3, JAR and BeWo cells for
175
studying placental endocrine function and gene expression [23,24,32–35,39,40]. Like normal trophoblast cells, BeWo, JAR and JEG-3 cell lines lack 17␣-hydroxylase/17,20 lyase activity but produce human chorionic gonadotropins, progesterone and aromatize exogenous androgens into estrogens. However, unlike normal trophoblast cells, these cell lines show only modest degrees of differentiation and secrete low amounts of hormones characteristic of syncytiotrophoblast, such as chorionic somatomammotropin [22,39,40]. Studies by Taylor et al. [41] have shown that BeWo cells have cytotrophoblast-like features and express low endogenous levels of aromatase. When cultured in presence of forskolin, a direct activator of adenylyl cyclase, these cells acquire an intermediate trophoblast phenotype with a ∼13.5-fold increase in aromatase activity. On the other hand, JEG-3 cells do not show any cytological changes in the presence of cAMP although there is an increase in aromatase activity [22]. We performed our initial studies in JEG-3 cells because of their ease of transfection, availability and rapid proliferation. However, we found that the JEG-3 cells exhibit aromatase activity that is approximately 1/50th that of primary cultures of human placental cells; CYP19 mRNA levels are undetectable by Northern blotting (Fig. 3) [36]. In consideration of this finding, we believe that the human trophoblast cells in primary culture provide a more physiologically relevant model system for mapping genomic regions required for placenta-specific expression of the human CYP19 gene than transfected human choriocarcinoma cells which express very low levels of aromatase.
Fig. 2. Aromatase activity (A) and Northern blot analysis of CYP19 mRNA in human trophoblasts before and after differentiation to syncytiotrophoblast in culture. (A) Freshly isolated cytotrophoblasts were suspended in DMEM containing 2% fetal bovine serum and plated at a density of 2 × 106 cells per dish. The aromatase activity (pmol androgen metabolized to estrogen/mg protein/min) was assayed over a 4-day-period by the incorporation of tritium from [1-3 H]androstenedione into water. The values are the mean ± S.E.M. of data from triplicate dishes of cells. (B) Total RNA (20 g per lane) obtained from freshly isolated cytotrophoblasts and syncytiotrophoblast after 24, 48 and 72 h in culture were analyzed for CYP19 mRNA by Northern blotting using a full-length 32 P-labeled human CYP19 cDNA probe [36].
176
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
Fig. 3. Aromatase activity (A) and Northern blot analysis of CYP19 mRNA (B) in human trophoblasts and JEG-3 cells. (A) Aromatase activity (pmol androgen metabolized to estrogen/mg protein/min) in trophoblast cells in primary culture and in JEG-3 cells was assayed by analysis of the incorporation of tritium from [1-3 H]androstenedione into water. Values shown are the mean ± S.E.M. of aromatase activity of triplicate dishes of cells on day 4 of culture. (B) Total RNA (20 g) obtained from freshly isolated cytotrophoblasts, syncytiotrophoblast after 4 days in culture and JEG-3 cells at 80–90% confluence was analyzed for CYP19 mRNA by Northern blotting using a full-length 32 P-labeled human CYP19 cDNA as probe [36].
1.5. Analysis of genomic sequences that mediate placenta-specific expression of CYP19 using cultured placental cells 1.5.1. Analysis of genomic sequences that mediate placenta-specific expression of CYP19 using primary cultures of human trophoblast cells To functionally map genomic sequences required for placenta-specific CYP19 gene expression, we have chosen to use primary cultures of human trophoblast cells which express high levels of aromatase. Fusion genes comprised of various amounts of CYP19 exon I.1 5 -flanking DNA, linked to the human growth hormone (hGH) structural gene, as reporter, were incorporated into a replication-defective human adenovirus vector (Ad5) and introduced into primary cultures of cytotrophoblasts and other cell types by infection. This results in highly efficient and reproducible transfer of fusion gene constructs into cells [36,42,43]. We observed that expression of CYP19I.1−2400 , CYP19I.1−923 and CYP19I.1−501 :hGH fusion genes increased in primary human trophoblasts in culture in concert with syncytiotrophoblast differentiation and induction of CYP19 gene expression. Expression of these fusion genes was undetectable in non-placental cells. On the other hand, we found that fusion genes containing 246, 201 and 125 bp of exon I.1 5 -flanking DNA were expressed in primary trophoblast cultures as well as in non-placental cells. By contrast, expression of the 42 bp containing fusion gene construct which includes only the minimal promoter region, remained essentially undetectable in placental and non-placental cells (Fig. 4). Taken together, all these results suggest that the region between −501 and −42 bp upstream from exon I.1 is sufficient for enhanced levels of CYP19 promoter I.1 activity in primary trophoblast cells in culture
and that the sequences between −501 and −246 bp contain silencer elements that may bind inhibitory transcription factors in non-placental cell types [36]. Further studies using deletion analysis, site-directed mutagenesis and electrophoretic mobility shift assays have indicated that two overlapping hexameric sequences (AGGTCA, −183 and −191 bp) which may bind a member(s) of the nuclear hormone receptor superfamily [35] and a G/C-rich sequence (−233 bp) which binds Sp1 [44] and other, as yet, unidentified transcription factors, are critical for maximum induction of CYP19I.1 promoter activity during syncytiotrophoblast differentiation [36]. 1.5.2. Analysis of genomic sequences that mediate placenta-specific expression of CYP19 using choriocarcinoma cell lines Toda et al. [32] have characterized two cis-acting elements by transient transfection of human BeWo choriocarcinoma cells, termed as hATRE-1 (human aromatase cytochrome P450 gene transcriptional regulatory element-1) and hATRE-2 which lie between −2238 and −2214 bp, and between −2141 and −2115 bp upstream of the transcriptional start site in exon I.1, respectively. Further characterization of these sites indicated that hATRE-1 acts as a repressor whereas the region between −2141 and −2115 bp acts as an enhancer, in response to 12-O-tetradecanoylphorbol-13-acetate and binds the transcription factor, C/EBP. Additionally, they have localized a cell-type specific enhancer element between −242 and −166 bp relative to the transcriptional start site in exon I.1 that is necessary for maximum transcriptional enhancement of aromatase in BeWo cells. Using JEG-3 choriocarcinoma cells to analyze the 5 -flanking region of CYP19 gene exon I.1, Yamada et al. [23] have reported that the initial 301 bp upstream of exon
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
177
Fig. 4. Expression of CYP19I.1:hGH fusion genes in primary human trophoblast cultures and in various cell lines. The different cell types shown were infected with 1 × 106 recombinant infectious viral particles containing CYP19I.1−2400 :hGH, CYP19I.1−923 :hGH, CYP19I.1−501 :hGH, CYP19I.1−246 :hGH, CYP19I.1−201 :hGH, CYP19I.1−125 :hGH and CYP19I.1−42 :fusion genes. Shown are the levels of hGH that accumulated in the medium over a 24-h-period between days 2 and 3 of culture. Values are the mean ± S.E.M. of data from 2–4 independent experiments, each conducted in triplicate [36].
I.1 may confer placenta-specific expression to the CYP19 gene. Recently, they have cloned a glial cells missing (GCM) motif protein that binds to a trophoblast-specific element within this region and may be involved in the regulation of CYP19 and possibly various other placenta-specific genes [34]. However, we have observed that the CYP19I.1:hGH fusion genes express at much lower levels in the JEG-3 cells than the primary cultures. Additionally, their pattern of expression is not similar to that seen in the primary trophoblast cells in culture [36]. Hence, although JEG-3 cells express endogenous aromatase and share many other characteristics of normal trophoblast cells, they may not provide a relevant system for delineating genomic sequences required for placenta-specific expression of the human CYP19 gene. 1.6. Use of transgenic mice to analyze regions of CYP19 gene required for placenta-specific expression Transgenic mice provide the most relevant model system to map genomic regions required for appropriate tissue-specific and developmental regulation of gene expression. Initially, we were uncertain whether the mouse would serve as an appropriate model system, since rodent placenta does not express aromatase. However, we observed that expression of CYP19I.1−2400 :hGH or CYP19I.1−501 :hGH fusion genes in transgenic mice was placenta-specific and developmentally regulated. These results clearly indicate that mouse placenta contains the necessary transcription factors to activate the human CYP19 gene placental-specific promoter but lacks the critical cis-acting elements required
for expression of the endogenous CYP19 gene. Similar studies on the ␣-subunit of the glycoprotein hormones, which is expressed in the placentae of only primates and horses [46,47], have indicated that lack of expression of the glycoprotein hormone ␣-subunit in rodent placenta was due to a single nucleotide change in a cyclic AMP-response element (CRE) and not due to the absence of essential regulatory transcription factors. However, unlike the ␣-subunit of the glycoprotein hormones, which appears to use a single promoter, the complexity of the CYP19 gene has made it difficult to define the genetic basis for placenta-specific expression of human aromatase by comparison of promoter sequences. Humans, cattle, horses and pigs, all appear to utilize placenta-specific promoters for tissue-specific expression of the CYP19 gene. However, in those species in which the 5 -flanking regions of the gene have been cloned, there is very low sequence identity and the placenta-specific promoters lie at variable distances upstream of the translation start site in exon II [48,49]. To localize CYP19I.1−501 :hGH fusion gene expression within the mouse placenta, in situ hybridization was performed using an antisense hGH probe. Reporter gene expression was absent in the spongiotrophoblast and the trophoblast giant cells, which produce many of the placental polypeptide hormones and steroidogenic enzymes [50,51]. On the other hand, the transgene was expressed as early as E10.5 specifically in the mouse labyrinthine trophoblast layer (Fig. 5.) [45]. This layer which is highly vascularized is trilaminar; its outer cells, which cover two layers of
178
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
Fig. 5. Expression of a CYP19I.1−501 :hGH fusion gene is restricted to labyrinthine trophoblast in transgenic mice. Placental tissues obtained from E10.5 or E17.5 fetal transgenic mice carrying the CYP19I.1−501 :hGH fusion gene (A–C) were processed for in situ hybridization by using a 35 S-labeled antisense hGH cRNA probe and exposed to photographic emulsion for 1–2 weeks. Bright and dark field microscopy then were performed. (A) Dark field micrograph of a placental tissue section from an E10.5 transgenic mouse carrying the CYP19I.1−501 :hGH fusion gene hybridized with radiolabeled hGH antisense cRNA probe. Silver grains indicating hGH mRNA transcripts appear as white spots. (B) Bright field micrograph of the haematoxylin-stained E10.5 placental tissue section shown in A. (C) Dark field micrograph of E17.5 placental tissue section incubated with radiolabeled hGH antisense cRNA probe. (D) Dark field micrograph of E17.5 placental section from a nontransgenic mouse incubated with radiolabeled hGH antisense cRNA probe (gc: trophoblast giant cell; sp: spongiotrophoblast; lab: labyrinthine trophoblast; ud: uterine decidua) [45].
syncytial cells, are bathed in maternal blood and thus play an important role in nutrient and gas exchange. In this respect, the labyrinthine layer of the mouse placenta, is most analogous to the human syncytiotrophoblast, which expresses aromatase and also is bathed in maternal blood [52]. These interesting findings suggest that 501 bp of exon I.1 5 -flanking sequence is sufficient for labyrinthine-specific expression of the human aromatase promoter in mouse.
2. Conclusions and future perspectives To functionally define genomic regions that direct appropriate cell-/tissue-specific and developmental expression of human CYP19 gene in placenta, studies in transgenic mice and primary cultures of human trophoblast cells were undertaken. Studies using primary cultures of human placental cells transfected with the CYP19I.1:hGH fusion genes suggest that the region between 501 and 42 bp upstream of exon I.1 contains response elements required for trophoblast-specific expression of the CYP19 gene, whereas
the region between −501 and −246 bp of exon I.1 may bind transcriptional repressors that prevent CYP19 gene expression in non-placental cells. In transgenic mice, a relatively small region (∼500 bp) >40,000 bp upstream of the protein coding region of the human CYP19 gene is able to direct expression in placenta in an appropriate cell-specific manner. Although, mouse placenta does not express aromatase, these findings indicate that transgenic mice can serve as an appropriate model system for defining regulatory regions surrounding the human CYP19 gene that mediate trophoblast-specific expression. These results also suggest that the transcription factors that direct placenta-specific expression of CYP19 are highly conserved among the animal species, although the response elements that bind these factors probably are not. Furthermore, the finding that transgene expression is confined to the labyrinthine trophoblast suggests that this region of the mouse placenta is most analogous to the human syncytiotrophoblast, which expresses high levels of aromatase. Studies are currently in progress to identify the critical cis-acting elements and characterize the transcription factors
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
that bind to these sites and direct appropriate expression of CYP19. This research, therefore, will provide insight into the molecular events that promote and maintain syncytiotrophoblast differentiation and culminate in placenta-specific expression of CYP19.
Acknowledgements This research was supported by NIH Grant 5 RO1 DK-31206. A.K. was supported in part by NIH Training Grant 5-T32-HD-07190-16.
References [1] C.R. Mendelson, E.E. Wright, J.C. Porter, C.T. Evans, E.R. Simpson, Preparation and characterization of polyclonal and monoclonal antibodies against human aromatase cytochrome P450 (P450arom) and their use in its purification, Arch. Biochem. Biophys. 243 (1985) 480–491. [2] S. Nakajin, M. Shimoda, P.F. Hall, Purification to homogeneity of aromatase from human placenta, Biochem. Biophys. Res. Commun. 134 (1986) 704–710. [3] J.T. Kellis Jr., L.E. Vickery, Purification and characterization of human placental aromatase cytochrome P450, J. Biol. Chem. 262 (1987) 4413–4420. [4] E.R. Simpson, Y. Zhao, V.R. Agarwal, M.D. Michael, S.E. Bulun, M.M. Hinshelwood, S. Graham-Lorence, T. Sun, C.R. Fisher, K. Qin, C.R. Mendelson, Aromatase expression in health and disease, Rec. Prog. Hormone Res. 52 (1997) 185–214. [5] E.R. Simpson, M.S. Mahendroo, G.D. Means, M.W. Kilgore, M.M. Hinshelwood, S. Graham-Lorence, B. Amarneh, Y. Ito, C.R. Fisher, M.D. Michael, et al., Aromatase cytochrome P450, the enzyme responsible for estrogen biosynthesis, Endocrine Rev. 15 (3) (1994) 342–355. [6] N. Fournet-Dulguerov, N.J. MacLusky, C.Z. Leranth, R. Todd, C.R. Mendelson, E.R. Simpson, F. F Naftolin, Immunohistochemical localization of aromatase cytochrome P450 and estradiol dehydrogenase in the syncytiotrophoblast of the human placenta, J. Clin. Endocrinol. Metabol. 65 (1987) 757–764. [7] K.P. McNatty, D.T. Baird, A. Bolton, P. Chambers, C.S. Corker, H. McLean, Concentration of oestrogens and androgens in human ovarian venous plasma and follicular fluid throughout the menstrual cycle, J. Endocrinol. 71 (1) (1976) 77–85. [8] S. Carreau, B. Biluska, J. Levallet, Male germ cells: a new source of estrogens in the mammalian testis, Annales d’Endocrinologie 59 (2) (1998) 79–92. [9] E.R. Simpson, G.E. Ackerman, M.E. Smith, C.R. Mendelson, Estrogen formation in stromal cells of adipose tissue of women: induction by glucocorticoids, Proc. Natl. Acad. Sci. U.S.A. 78 (1981) 5690–5694. [10] J.M. Grodin, P.K. Siiteri, P.C. MacDonald, Source of estrogen production in postmenopausal women, J. Clin. Endocrinol. Metabol. 36 (1973) 207–214. [11] F. Naftolin, K.J. Ryan, I.J. Davies, V.V. Reddy, F. Flores, Z. Petro, M. Kuhn, R.J. White, Y. Takaoka, L. Wolin, The formation of estrogens by central neuroendocrine tissues, Recent Prog. Hormone Res. 31 (1975) 295–319. [12] S. Diano, T.L. Horvath, G. Mor, T. Register, M. Adams, N. Harada, F. Naftolin, Aromatase and estrogen receptor immunoreactivity in the coronory arteries of monkeys and human subjects, Menopause 6 (1) (1999) 21–28.
179
[13] K. Toda, E.R. Simpson, C.R. Mendelson, Y. Shizuta, M.W. Kilgore, Expression of the gene encoding aromatase cytochrome P450 (CYP19) in fetal tissues, Mol. Endocrinol. 8 (1994) 210–217. [14] K.J. Doody, B.R. Carr, Aromatase in human fetal tissues, Am. J. Obstet. Gynecol. 10 (1984) 241–244. [15] T. Price, J. Aitkin, E.R. Simpson, Relative expression of aromatase cytochrome P450 in human fetal tissues as determined by competitive polymerase chain reaction (PCR) amplification, J. Clin. Endocrinol. Metabol. 74 (1992) 879–883. [16] R.B. Everett, P.C. MacDonald, Endocrinology of the placenta, Ann. Rev. Med. 30 (1979) 473–488. [17] M.P. Steinkampf, C.R. Mendelson, E.R. Simpson, Regulation by FSH of the synthesis of aromatase cytochrome P-450 (P450arom) in human granulosa cells, Mol. Endo. 1 (1987) 465–471. [18] Y. Zhao, J.E. Nichols, S.E. Bulun, C.R. Mendelson, E.R. Simpson, Aromatase P450 gene expression in human adipose tissue. Role of a Jak/STAT pathway in regulation of adipose-specific promoter activity, J. Biol. Chem. 270 (1995) 16449–16457. [19] M.J. Lanoux, W.H. Cleland, C.R. Mendelson, B.R. Carr, E.R. Simpson, Factors affecting the conversion of androstenedione to estrogens by human fetal hepatocytes in primary monolayer culture, Endocrinology 117 (1985) 361–368. [20] J.E. Nestler, Modulation of aromatase and P450 cholesterol side-chain cleavage enzyme activities in human placental cytotrophoblasts by insulin and insulin-like growth factor I, Endocrinology 121 (5) (1987) 1845–1852. [21] J. Lobo, F.L. Bellino, Estrogen synthetase (aromatase) activity in primary term placental cells: effects of cell preparation, growth medium and serum on adenosine 3 , 5 -monophosphate response, J. Clin. Enocrinol. Metabol. 69 (1989) 868–874. [22] G.E. Ringler, J.F. Strauss, In vitro systems for the study of human placental endocrine function, Endocrine Rev. 11 (1990) 105–123. [23] K. Yamada, N. Harada, S. Honda, Y. Takagi, Regulation of placenta-specific expression of the aromatase cytochrome P450 gene, J. Biol. Chem. 270 (1995) 25064–25069. [24] K. Toda, M. Terashima, T. Kawamoto, H. Sumimoto, Y. Yokoyama, I. Kuribayashi, Y. Mitsuuchi, T. Maeda, Y. Yamamoto, Y. Sagara, H. Ikeda, Y. Shizuta, Structural and functional characterization of human aromatase P450 gene, Eur. J. Biochem. 193 (1990) 559–565. [25] N. Harada, K. Yamada, K. Saito, N. Kibe, S. Dohmae, Y. Takagi, Structural characterization of the human estrogen synthetase (aromatase) gene, Biochem. Biophys. Res. Commun. 166 (1990) 365–372. [26] G.D. Means, M.S. Mahendroo, J.C. Corbin, J.M. Mathis, F.E. Powell, C.R. Mendelson, E.R. Simpson, Structural analysis of the gene encoding human aromatase cytochrome P-450, the enzyme responsible for estrogen biosynthesis, J. Biol. Chem. 264 (1989) 19385–19391. [27] S. Chen, M.J. Besman, R.S. Sparkes, S. Zollman, I. Klisak, T. Mohandes, P.F. Hall, J.E. Shively, Human aromatase: cDNA cloning, southern blot analysis and assignment of the gene to chromosome 15, DNA 7 (1988) 27–38. [28] M.S. Mahendroo, C.R. Mendelson, E.R. Simpson, Tissue-specific and hormonally-controlled alternative promoters regulate aromatase cytochrome P450 gene expression in human adipose tissue, J. Biol. Chem. 268 (1993) 19463–19470. [29] Y. Zhao, C.R. Mendelson, E.R. Simpson, Characterization of the sequences of the human CYP19 (aromatase) gene that mediate regulation by glucocorticoids in adipose stromal cells and fetal hepatocytes, Mol. Endocrinol. 9 (1995) 340–349. [30] M.S. Mahendroo, G.D. Means, C.R. Mendelson, E.R. Simpson, Tissue-specific expression of human P450arom: the promoter responsible for expression in adipose is different from that utilized in placenta, J. Biol. Chem. 266 (1991) 11276–11281. [31] G.D. Means, M.W. Kilgore, M.S. Mahendroo, C.R. Mendelson, E.R. Simpson, Tissue-specific promoters regulate aromatase cytochrome P450 gene expression in human ovary and fetal tissues, Mol. Endocrinol. 5 (1991) 2005–2013.
180
A. Kamat, C.R. Mendelson / Journal of Steroid Biochemistry & Molecular Biology 79 (2001) 173–180
[32] K. Toda, Y. Shizuta, Identification and characterization of cis-acting regulatory elements for the expression of the human aromatase cytochrome P-450 gene, J. Biol. Chem. 269 (1994) 8099–8107. [33] K. Toda, L.X. Yang, Y. Shizuta, Transcriptional regulation of the human aromatase cytochrome P450 gene expression in human placental cells, J. Steroid. Biochem. Mol. Biol. 53 (1995) 181–190. [34] K. Yamada, S. Honda, N. Harada, T. Okazaki, A GCM motif protein is involved in placenta-specific expression of human aromatase gene, J. Biol. Chem. 274 (1999) 32279–32286. [35] T. Sun, Y. Zhao, D.J. Mangelsdorf, E.R. Simpson, Characterization of a region upstream of exon I.1 of the human CYP19 (aromatase) gene that mediates regulation by retinoids in human choriocarcinoma cells, Endocrinology 139 (1998) 1684–1691. [36] A. Kamat, J.L. Alcorn, C. Kunczt, C.R. Mendelson, Characterization of the regulatory regions of the human aromatase (P450arom) gene involved in placenta-specific expression, Mol. Endocrinol. 12 (1998) 1764–1777. [37] H.J. Kliman, J.E. Nestler, E. Sermasi, J.M. Sanger, J.F. Strauss, Purification, characterization and in vitro differentiation of cytotrophoblasts from human term placentae, Endocrinology 118 (1986) 1567–1582. [38] J.O. Lobo, F.L. Bellino, L. Bankert, Estrogen synthetase activity in human term placental cells in monolayer culture, Endocrinology 116 (1985) 889–895. [39] H.-P. Hohn, M. Linke, B. Ugele, H.-W. Denker, Differentiation markers and invasiveness: discordent regulation in normal trophoblast and choriocarcinoma cells, Exp. Cell Res. 244 (1998) 249–258. [40] R.D. Bahn, A. Worsham, K.V. Speeg, M. Ascoli, D. Rabin, Characterization of steroid production in cultured human choriocarcinoma cells, J. Clin. Endocrinol. Metabol. 52 (1981) 447–450. [41] R.N. Taylor, E.D. Newman, S. Chen, Forskolin and methotrexate induce an intermediate trophoblast phenotype in cultured human choriocarcinoma cells, Am. J. Obstet. Gynecol. 164 (1991) 204–210. [42] B.L. Davidson, E.D. Allen, K.F. Kozarsky, J.M. Wilson, J. Roessler, A model system for in vivo gene transfer into the central nervous system using an adenoviral vector, Nat. Genet. 3 (1993) 219–223. [43] J.L. Alcorn, E. Gao, Q. Chen, M.E. Smith, R.D. Gerard, C.R. Mendelson, Genomic elements involved in transcriptional regulation
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
of the rabbit surfactant protein-A gene, Mol. Endocrinol. 7 (1993) 1072–1085. M.R. Briggs, J.T. Kadonaga, S.P. Bell, R. Tijan, Purification and biochemical characterization of the promoter-specific transcription factor Sp1, Science 234 (1986) 47–52. A. Kamat, K.H. Graves, M.E. Smith, J.A. Richardson, C.R. Mendelson, A 500 bp region, ∼40 kb upstream of the human CYP19 (aromatase) gene, mediates placenta-specific expression in transgenic mice, Proc. Natl. Acad. Sci. U.S.A. 96 (1999) 4575–4580. J.A. Bokar, R.A. Keri, T.A. Farmerie, R.A. Fenstermaker, B.A. Andersen, D.L. Hamernik, J. Yun, T. Wagner, J.H. Nilson, Expression of the glycoprotein hormone ␣-subunit gene in the placenta requires a functional cyclic AMP response element, whereas a different cis-acting element mediates pituitary-specific expression, Mol. Cell Biol. 9 (11) (1989) 5113–5122. J.H. Nilson, J.A. Bokar, C.M. Clay, T.A. Farmerie, R.A. Fenstermaker, D.L. Hamernik, R.A. Keri, Different combinations of regulatory elements may explain why placenta-specific expression of the glycoprotein hormone ␣-subunit gene occurs only in primates and horses, Biol. Reprod. 44 (1991) 231–237. M.M. Hinshelwood, Z. Liu, A.J. Conley, E.R. Simpson, Demonstration of tissue-specific promoters in non-primate species that express aromatase P450 in placentae, Biol. Reprod. 53 (5) (1995) 1151–1159. I. Choi, D.L. Troyer, D.L. Cornwell, K.R. Kirby-Dobbels, W.R. Collante, F.A. Simmen, Closely related genes encode developmental and tissue isoforms of porcine cytochrome P450 aromatase, DNA Cell Biol. 16 (1997) 769–777. M.J. Soares, B.M. Chapman, C.A. Rasmussen, G. Dai, T. Kamei, K.E. Orwig, Differentiation of trophoblast endocrine cells, Placenta 17 (1996) 277–289. S.T. Suhr, J.O. Rahal, K.E. Mayo, Mouse growth-hormone releasing hormone: precursor structure and expression in brain and placenta, Mol. Endocrinol. 3 (1989) 1693–1700. W.D. Billington, Organization, ultrastructure and histochemistry of the placenta: immunological considerations, in: R.G. Edwards, C.W.S. Howe, M.H. Johnson (Eds.), Immunobiology of Trophoblast, Cambridge University Press, Cambridge, 1975, pp. 67–85.