Abstract
Expression analysis by reverse transcriptase (RT)-PCR indicates that human adipose tissue is not likely to perform de novo synthesis of steroid hormones from cholesterol because the mRNAs of cytochromes P450scc and P450c17, and of the steroidogenic-related proteins, steroidogenic acute regulatory protein and steroidogenic factor 1, were not detected. Instead, our data support an intracrine role of adipose tissue, in which adrenal dehydroepiandrosterone sulfate (DHEA-S), the most abundant circulating androgen in man, is selectively uptaken, desulfated, and converted into bioactive androgens and estrogens. Three organic anion-transporting polypeptides-B, -D, and -E, presumably involved in DHEA-S transmembrane transport, were demonstrated at the mRNA level. While sulfotransferase expression was not found, the occurrence of steroid sulfatase (STS), converting DHEA-S to DHEA, was established at the mRNA, protein and catalytic activity levels. The 5′-rapid amplification of cDNA ends analysis showed that STS transcription in adipose tissue is regulated by the use of two promoters which differ from the prevalent placental one. The adipose transcripts contain a distinct untranslated first exon, 0a or 0b, followed by a common partially translated exon 1b, and nine other exons that are also shared by the main placental transcript. The presence of an upstream open reading frame in the new transcript variants could lead to an N-terminal divergence restricted to the cleavable signal peptide and thus not interfering with the catalytic activity of the mature STS protein. The adipose transcripts are also present in the placenta as minor isoforms. Western blotting revealed the characteristic ~64 kDa band of STS in both the placenta and adipose tissue. The specific enzymatic activity of STS in adipocytes was 118 pmol/106 cells per hour, about 50–100 times lower than in the placenta. A similar rate of [3H] DHEA-S uptake plus desulfation was measured in preadipo-cytes and adipocytes, equivalent to 40–45 pmol/106 cells per hour. Thus, an excessive accumulation of fat may out-compete other peripheral organs that are also dependent on intracrine DHEA-S utilization, especially when the adrenal production is low or declining with aging.
Introduction
Adipose tissue is no longer viewed merely as the site of diet-induced storage of excess energy, controlled by external lipogenic and lipolytic hormones, but is also recognized as a bona fide endocrine organ releasing adipokines and other bioactive peptides influencing appetite, energy homeostasis, and reproductive competence (Frayn et al. 2003, Kershaw & Flier 2004).
This tissue also expresses several terminal steroid-transforming enzymes leading to the secretion of active steroid hormones such as 3β-hydroxysteroid dehydrogenase (3β-HSD)/Δ5-Δ4-isomerase, 17β-HSD, and 5α-reductase for androgens, cytochrome P450arom for estrogens, and 11β-HSD type 1 for glucocorticoids (reviewed by Belanger et al. 2002). Given the proportional mass of fat depots, their secretions may account for up to 100% of circulating estrogens in postmenopausal women and 50% of circulating testosterone in premenopausal women (Kershaw & Flier 2004). Expression of cytochrome P450c17 and conversion of progesterone to 17α-hydroxyprogesterone has been reported in adipose tissue (Puche et al. 2002), but so far, there is no evidence of other key enzymes for progestogen and corticoid synthesis from cholesterol.
In humans and other primates, the adipose tissue may be supplied with preformed steroids by the adrenal glands. These secrete not only corticoids, but also dehydroepiandrosterone (DHEA) and much greater amounts of DHEA sulfate (DHEA-S), which can be converted into potent androgens and/or estrogens by peripheral tissues, a process defined as intracrine (Labrie et al. 2001). DHEA-S is negatively charged at physiological pH and its uptake requires an active transport across the plasma membrane, which is performed by members of two organic anion carrier protein families: the organic anion-transporting polypeptide (OATP) family and the organic anion transporter (OAT) family. These transporters are expressed in several tissues and display wide and overlapping substrate specificities for many organic compounds, including steroid conjugates. In particular, several OATPs can transport sulfate conjugates of steroids, such as OATP-A, OATP-B, OATP-C, OATP-8, and OATP-E (Hagenbuch & Meier 2003), whereas in the OAT family, only OAT-3 and OAT-4 can mediate the cellular uptake of certain steroid conjugates in the kidney, liver, brain, and placenta (Cha et al. 2000, 2001).
Once internalized, DHEA-S must undergo hydrolysis of the sulfate group, carried out by the enzyme steroid sulfatase (STS), before being transformed further. STS is a membrane-bound microsomal enzyme, ubiquitously expressed in mammalian tissues (Martel et al. 1994), that hydrolyzes various alkyl (e.g. DHEA-S) and aryl steroid sulfates (e.g. estrone sulfate) (Reed et al. 2005). In man, the STS gene contains ten exons spread over 146 kbp in the short arm of the X chromosome (Yen et al. 1988). The deduced amino acid sequence consists of 583 amino acids with an N-terminal signal peptide of 21or 23 residues, as there are two dipeptides in tandem that equally fulfill the requirements for recognition by the signal peptidase (Stein et al. 1989).
In the present study on human adipose tissue, we have examined by reverse transcriptase (RT)-PCR its steroidogenic competence, the expression of organic anion transporters belonging to the OATP and OAT families, the transmembrane transport of DHEA-S, the transcriptional control of the gene encoding STS, and the specific activity of this enzyme.
Materials and Methods
Collection of adipose tissue for RNA and protein analyses
Biopsies of subcutaneous abdominal adipose tissue were obtained from six female (age: 46–61 years) and three male (age: 39–62 years) patients undergoing abdominal surgery for nonmalignant conditions and were kept at −80 °C until RNA or protein analyses. Samples of fresh adipose tissue were also collected and transported immediately to the laboratory for STS assays. All tissue samples were used with full patient consent and institutional review board approval.
RNA extraction and expression analyses
Total RNA was extracted with Trizol reagent (Invitrogen) and kept at −80 °C until use. Reverse transcription of 2 μg of total RNA was performed in 20 μl final volume with random hexamers and the ThermoScript RT-PCR System Kit (Invitrogen). After incubation for 10 min at 25 °C, RT was carried out for 30 min at 55 °C, followed by RT termination for 5 min at 85 °C. Total single-stranded cDNA was amplified by PCR in a 25 μl reaction volume containing 10 mM Tris–HCl (pH 8.3), 50 mM KCl, 1.5–2 mM MgCl2, 0.2 mM of each dNTP, 0.2 μM of the respective primers and 1.25 U of Biotherm Taq DNA polymerase (Società Italiana Chimici, Rome, Italy). Oligonucleotide primers are listed in Table 1 and amplification conditions in Table 2. The extension phase of the last cycle was prolonged by 10 min. An aliquot of the products was run on 1% agarose gel. For each PCR, a negative control was prepared by replacing the cDNA solution with sterile water to check for cross contamination.
Determination of the transcription start site (TSS) of the STS gene expressed in the adipose tissue
The TSS of the STS gene was determined with the method of RNA ligase-mediated rapid amplification of cDNA 5′-ends (RLM-5′-RACE) using the FirstChoice RLM-RACE kit (Ambion, Celbio, Milan, Italy). Briefly, 10 μg of adipose tissue total RNA was treated with calf intestinal phosphatase to remove the 5′-phosphate from truncated RNAs, leaving a 5′-OH end. Total RNA was then treated with tobacco acid pyrophosphatase to remove the 5′-cap from full-length mRNAs, leaving a 5′-phosphate to which a 5′-RACE RNA adapter oligonucleotide was ligated with T4 RNA ligase. Ligated mRNAs were then reverse-transcribed with random decamers.
STS transcripts were PCR-amplified using the 5′-RACE outer primer and a specific 3′-reverse primer (STS-4) designed on the exon 5 of the STS gene (Stein et al. 1989). The amplification procedure consisted of 2 min at 95 °C, followed by a touchdown PCR with annealing temperatures decreasing from 68 °C to 56 °C over 12 cycles and the final 28 cycles at 56 °C. The extension phase of the last cycle was prolonged by 10 min. The diluted products were subjected to second and third rounds of amplification using the 5′-RACE inner primer and two specific 3′-reverse primers (STS-6 and STS-10) selected on exon 3 and exon 2, respectively.
The amplicon from the last PCR was purified from the sliced gel band and ligated into a pGEM-T vector using a pGEM-T Vector System I (Promega). Plasmids from positive colonies were purified and four clones were sequenced.
3′-RACE analysis
For 3′-RACE analysis, 2 μg of adipose tissue total RNA was incubated in a 20 μl reaction volume with 200 U of SuperScript II RT (Invitrogen), 0.5 mM of dNTPs, 10 mM dithiothreitol (DTT) and 0.5 μM of dT17 primer (5′-GAC TCG AGT CGA CAT CGA TTT TTT TTT TTT TTT TT-3′) for 1 h at 50 °C. The first-strand mixture was diluted fivefold and 2 μl were added to 50 μl of the PCR buffer containing 0.2 mM of dNTPs, 0.2 μM of the anchor primer (5′-CTG GTT CGG CCC AGA CTC GAG TCG ACA TCG-3′), 2.5 U of Biotherm Taq DNA polymerase, and the gene-specific STS-5 primer. The first-round PCR product was diluted 1:100 and used as a template for the second round of PCR using the STS-7 and anchor primers. The amplification procedure with touchdown PCR was as above. The resultant amplicon was purified from the sliced gel band and directly sequenced.
Nucleotide sequencing
Sequencing was performed on double-stranded DNA directly from PCR products or after cloning using the ABI PRISM dye terminator cycle sequencing core kit (Applied Biosystems, Monza, Italy). Electrophoresis of sequencing reactions was completed on the ABI PRISM model 377, version 2.1.1 automated sequencer. The homology searches were carried out using the Basic Blast program version 2.0 (http://www.ncbi.nlm.nih.gov/BLAST/), and alignments were performed using the ClustalW program (http://www2.ebi.ac.uk/clustalw/).
Western blot analysis of adipose tissue STS
Samples of adipose tissue were homogenized in an Ultra-Turrax at 4 °C in 100 mM potassium phosphate buffer, pH 7.4, containing 250 mM sucrose, 4 mM MgCl2, 1 mM EDTA, 0.5 mM phenylmethylsulfonylfluoride, and 2 μg/ml each of leupeptin, pepstatin A and aprotin. Cell-free supernatants were obtained after three centrifugations at 1000 g for 15 min at 4 °C.
Western immunoblot analysis of STS was done with a specific rabbit antiserum against human STS kindly provided by Dr Bernhard Ugele (Frauenklinik-Innenstadt, University Hospital Munich, Munich, Germany). Human placenta was used as a positive control tissue. Proteins from adipose tissue (50 μg) and placenta (5 and 0.5 μg) were size-fractioned onto 10% SDS-PAGE (Nupage, Invitrogen) and transferred onto a nitrocellulose filter (Roche Applied Science). After a 2 h blocking step at 4 °C with 5% BSA in TBS-T buffer (20 mM Tris–HCl, pH 7.6, 137 mM NaCl and 0.2% Tween-20), the filter was hybridized with 1/10 000 dilution of the antibody against human STS in TBS-T buffer containing 5% BSA overnight at 4 °C. The filter was then washed (3×5 min; 1× 15 min) with TBS-T and further incubated for 1 h at room temperature with 1:5000 dilution of goat horseradish peroxidase (HRP)-conjugated antirabbit IgG antibody (Pierce, Celbio, Milan, Italy) in TBS-T at 25 °C. After washing as above, HRP activity was detected using the SuperSignal West-Dura chemiluminescent kit (Pierce).
Sulfatase assay
Tissues were homogenized with an Ultra-Turrax in a buffered medium (100 mM KCl, 16 mM K2HPO4, 4 mM KH2PO4, 1 mM DTT, 1 mM EDTA, 4 mM nicotinamide) and homogenates were centrifuged four times at 1000 g for 15 min at 4 °C. In one experiment, adipocytes were homogenized by flushing through a syringe needle. Supernatants were preincubated for 20 min at 37 °C and then incubated in 1 ml final volume with 74 kBq of [1,2,6,7-3H] DHEA-S (25 μM), previously dissolved in 10 μl of propylene glycol, for 15–120 min at 37 °C. Boiled homogenates were incubated as controls. All incubations were performed in duplicate or triplicate.
Enzyme activity was terminated by adding water-saturated ethyl acetate. [14C]DHEA was then added as a tracer for recovery control and radiochemical identification. Free steroids were extracted four times with 2 vol of ethyl acetate and separated by thin layer chromatography (plastic sheets precoated with silica gel 60 F254; Merck, Milan, Italy) with the solvent systems cyclohexane/ethyl acetate (95:5, v/v; 5–7 runs) for defatting, and cyclohexane/ethyl acetate (1:1, v/v; 2 runs). Chromatoplates were autoradiographed for 5 days and the spots corresponding to unconjugated DHEA were removed, eluted in 20 ml of acetone and counted in a Packard Tri-Carb 1500 Liquid Scintillation Analyzer (Perkin Elmer, Monza, Italy), with a dual-label program.
For each incubate, conclusive identification of labeled metabolites with tracer was based on derivatization by acetylation and rechromatography. Percent conversions were calculated from the last [3H/14C] ratio in the purification procedure and were thus corrected for losses.
Isolation, culture, and differentiation of human preadipocytes
For preadipocyte culture, adipose tissue samples were collected by liposuction from five healthy female patients (age: 25–40 years) undergoing plastic reduction procedure. After trimming of the surrounding fibrous tissue, dissected adipose lobules were rinsed four times with PBS containing 1% antibiotic and antimycotic solution (AF, Sigma), minced and digested with type IA collagenase (1 mg/ml) (Sigma) for 45 min at 37 °C. Disaggregated cells were passed through a 80 μm nylon mesh to remove undigested tissue and centrifuged at 400 g for 5 min. They were then cultured at the density of 2×105 cells/cm2 in DMEM–LG medium (Sigma) containing 10% fetal calf serum (FCS, Biochrom-Seromed, Berlin, Germany), 0.1% AF solution, and 0.1 μg/ml bovine fibroblast growth factor. Two days after reaching maximum confluence, cells were induced to differentiate into mature adipocytes by treatment for 10 days with DMEM–LG medium containing 10% FCS, 10 μg/ml insulin, 10−3 M 3-isobutyl-1-methylxanthine (IBMX), 6×10−5 M indomethacin, and 10−6 M dexamethasone (DEX) (Sigma).
[3H]DHEA-S transport studies along with the sulfatase assay in adipose cells
Cultured preadipocytes and adipocytes were centrifuged and resuspended to 1–2×106 cells/ml in ice-cold transport buffer containing 142 mM NaCl, 5 mM KCl, 1.2 mM MgSO4, 1.5 mM CaCl2, 1 mM KH2PO4, 5 mM glucose, 12.5 mM HEPES, and pH 7.4. For transport studies, aliquots of 0.9 ml of cell suspensions were preincubated for 20 min at 37 °C. The experiment was started by adding 100 μl of transport buffer containing 74 kBq of [1,2,6,7-3H]DHEA-S (50 μM). At fixed times, the reaction was terminated by adding water-saturated ethyl acetate. Since [14C]DHEA-S to be used as an internal tracer is not commercially available, we have measured conversion to DHEA as an estimate of both transmembrane transport and sulfatase activity. Hence, [14C]DHEA was added as a tracer for recovery control and identification and free DHEA was measured as above. [14C]-labeled androstenedione, testosterone, and androstene-3β,17β-diol were also added as tracers in the 6 h incubates.
Results
Expression analysis of steroidogenic enzymes and related proteins in adipose tissue
To establish the steroidogenic competence of human adipose tissue, we analyzed by RT-PCR with specific primers (Table 1) the expression of cytochromes P450scc, P450c17, and P450arom, along with the expression of two steroidogenesis-related proteins, such as steroidogenic factor 1 (SF-1) and steroidogenic acute regulatory protein (StAR). RNAs extracted from human adrenal cortex and placenta were used as controls. Cytochrome P450scc mRNA was expressed in both the adrenal cortex and placenta, P450c17 mRNA only in the adrenal and that of P450arom only in the placenta. In the adipose tissue, P450scc and P450c17 mRNAs were not detected, whereas a positive signal was obtained for P450arom mRNA, though at a lower level than in the placenta (Fig. 1).
As expected, StAR and SF-1 mRNAs were observed in the adrenal cortex, but not in the placenta, which lacks both proteins (Sugawara et al. 1995, Strauss et al. 1996). Adipose tissue was also negative for both transcripts (Fig. 1), confirming a previous report on the absence of SF-1 in mammary fat (Clyne et al. 2002).
According to the literature, dehydroepiandrosterone sulfotransferase mRNA is expressed in the adrenal cortex, but not in the placenta (Luu-The et al. 1995), while estrogen sulfotransferase mRNA is found in both tissues (Miki et al. 2002). None of these enzyme mRNAs showed a detectable expression in the human adipose tissue (Fig. 1), whereas STS mRNA was expressed in both control and adipose tissues. When five samples of cultured preadipocytes, two samples of cultured adipocytes, and nine samples of bioptic adipose tissue were similarly analyzed by RT-PCR, all were found to be positive for STS mRNA.
Expression analysis of known members of the OATP and OAT gene superfamilies in adipose tissue
The expression of the known members of the OATP (subtypes A, B, C, D, E, and 8) and OAT (subtypes 3 and 4) gene superfamilies that might transport sulfate steroids across membranes was assessed by RT-PCR with specific sets of primers in adipose tissue. As shown in three representative tissue samples in Fig. 2, OATP-B, OATP-D, and OATP-E mRNAs were constitutively expressed, whereas the mRNAs of OATP-A, OATP-C, OATP-8, OAT-3, and OAT-4 were not detected in any sample. The identity of the amplificates was verified by sequencing.
As OATP-B’s ability to transport DHEA-S is well established in the literature, its expression was confirmed by RT-PCR in all samples of adipose tissue and in cultured preadipocytes and adipocytes.
RLM-5′-RACE and 3′-RACE of STS expressed in adipose tissue
The expression of STS mRNA was initially analyzed by RT-PCR using three sets of primers covering the whole coding region (STS-3–STS-4; STS-1–STS-2; STS-5–STS-8). Using the first set, in which the forward primer is located immediately upstream of the initiation codon of the placental STS cDNA sequence (Stein et al. 1989), we could not amplify the first fragment of the coding region. However, two single fragments of the expected sizes were consistently detected on agarose gel with the other two sets of primers covering the remaining coding region. Sequencing in both directions of these fragments showed complete identity between the adipose amplificates and the placental cDNA sequence. These results confirmed the STS expression in adipocytes, but hinted at a sequence divergence at the 5′ terminus.
The possible occurrence of alternative splicing in adipose tissue was probed by RLM-5′-RACE, using the placenta as a control. Unlike conventional RACE, which amplifies also truncated RNAs, RLM-5′-RACE amplifies only capped full-length RNAs. Two transcripts of the STS gene were found in adipose tissue, as compared to a single transcript in the placenta. The sequencing of the placental transcript confirmed the presence of a first exon (here called exon 1a) that is partially translated into the first 3 aa of the placental protein. The TSS was found 205 nt upstream of the first ATG. The ATG context matches only partially the proposed consensus sequence for the initiation of translation (Kozak 1996), because it lacks a G base at position +4.
As shown in Fig. 3, the two adipose transcripts contain different untranslated first exons (here named 0a and 0b) that are spliced to a common exon 1b. This is then spliced to exon 2, which is identical to that found in the placenta. Exon 0a is 100 nt long and lies 43.28 kb upstream of exon 1b, whereas exon 0b is 77 nt long and lies 42.8 kb upstream of the same exon. The cDNA sequences of the two adipose transcripts have been submitted to the EMBL Nucleotide Sequence Database under the accession numbers AM072429 and AM072428, respectively. The introns display splice signals consistent with the GT/AG rule, except the first intron after exon 0a, in which the 5′ splice donor is GC instead of GT. However, this represents the major splice variant (Thanaraj & Clark 2001).
The common exon 1b is 129 nt long and lies 62.19 kb upstream of exon 2. After a segment of 97 nt downstream of its 5′ end, exon 1b contains an ATG with a context matching the proposed consensus sequence (Kozak 1996). As this ATG is in frame with the coding sequence of exon 2, the protein encoded in adipose tissue will be 7 aa longer than the placental protein, whose first four aa differs from aa 8–11 of adipose STS (Fig. 4).
To confirm these results, RT-PCR analyses were performed with three different sets of primers, in which a common reverse primer was designed on exon 2 and the sense primers on adipose exons 0a and 0b and placental exon 1a, respectively. As shown in Fig. 5, the adipose tissue does not use the placental promoter, whereas low levels of adipose-type transcripts were found in the placenta. Although analyses were not quantitative, the signals with the set of primers specific for exon 0b were consistently more intense than those for exon 0a.
With regard to the 3′-UTR, 3′-RACE analysis provided a single amplification fragment whose sequence and polyadenylation signal, located 428 bp after the stop codon, were found to be identical to those reported by Stein and co-workers (1989).
Western blotting analysis
The presence of the STS protein was assessed in the samples of adipose tissue and placenta as a control by immunoblot analysis with a polyclonal antibody against human STS. As illustrated in Fig. 6, a protein band was observed in both placenta and adipose tissue samples. In the latter, the intensity of the immunoreaction was much lower than in the placentas, considering that only 0.5 (Pl1) or 1 (Pl2) μg of placental proteins were loaded instead of 50 μg, as in the adipose tissue preparations. This result is in agreement with the biochemical analysis showing that STS activity in adipose tissue is about 1/50 to 1/100 of that measured in placenta (see below). In addition to the characteristic ~64 kDa band of STS, a supplementary band of ~100 kDa was detected in both adipose and placental samples, which is likely due to nonspecific binding.
An attempt to localize STS in adipose tissue by immunohistochemistry with the same antibody was inconclusive, probably because the low level of expression of this enzyme in adipocytes is not detectable in their scanty amount of cytoplasm. Conversely, a strong cytoplasmic labeling was observed in placental syncytiotrophoblast (data not shown).
Measurement of STS activity
STS activity was determined by measuring the conversion of [3H]DHEA-S to DHEA. Homogenates of two adipose tissue samples as well as one sample of homogenized cultured adipocytes were incubated with 25 μM [3H]DHEA-S, a saturating concentration since the Km values reported for human STS were 9.59 μM for the purified protein (Hernandez-Guzman et al. 2001), 7.8 μM (Suzuki et al. 1992) and 16 μM (Gibb & Lavoie 1984). All incubations were carried out with 1 or 2 mg total protein for 30, 60 and 90 min, because an incubation time of 15 min was preliminarily found sufficient to get a significant signal (0.45% of conversion to DHEA) with 1 mg protein. Under these conditions, the rate of DHEA-S hydrolysis was linear with the incubation time (r>0.94). A homogenate of placental tissue was incubated as a positive control, but for shorter incubation times and with less protein (Table 3). The percent conversions with boiled tissues were always subtracted from experimental values to compensate for the spontaneous hydrolysis of the conjugated substrate.
The specific enzymatic activity of STS in cultured adipocytes was 118 pmol/106 cells per hour, in adipose tissue 1 and 2, 80 and 140 pmol/mg per hour respectively, that is about 50–100 times lower than in the placenta 7 nmol/mg per hour (Table 3).
[3H]DHEA-S transport studies along with the sulfatase assay in adipose cells
To measure [3H]DHEA-S transmembrane transport and intracellular hydrolysis, aliquots of suspensions of preadipo-cytes and adipocytes were incubated with the conjugate, whose final concentration was adjusted to 50 μM with unlabeled DHEA-S, since a Km of 26 μM for [35S]DHEA-S uptake was reported in isolated mononucleated trophoblastic cells (Ugele et al. 2003). After assessing that a minimum incubation time of 1 h was necessary to obtain a significant signal, preadipocytes were incubated for 3 and 6 h, while adipocytes were incubated for 6 h only. Incubations were performed singly, owing to the low numbers of cells available. Free DHEA was measured as above.
Adipocytes produced free DHEA at the rate of 45.83 pmol/106 cells per hour, while preadipocyte rates were 40.8 and 43.3 pmol/106 cells per hour. A search for further metabolites of DHEA, such as androstenedione, testosterone, and androst-5-ene-3β,17β-diol, gave negative results.
Discussion
The present study indicates that de novo biosynthesis of progestogens and androgens from cholesterol is unlikely in human adipose tissue, because the mRNAs of key steroidogenic enzymes, such as cytochromes P450scc and P45c17, and of proteins specifically expressed in steroidogenic cells, such as StAR and SF-1, could not be demonstrated in any of the samples. The failure to amplify cytochrome P450c17 cDNA, even with two other sets of primers (data not shown), is at variance with the report by Puche and co-workers (2002), but can be explained by the fact that they used not only RT-PCR, but also Southern blotting analysis to intensify the signal. The level of expression must be minimal anyway.
Our data support an intracrine role of adipose tissue as a terminal activator of circulating inactive androgen precursors into potent sex steroids. The detection of P450arom mRNA confirms previous findings about the occurrence of aromatizing activity for estrogen synthesis in adipose tissue, as reviewed by Belanger et al.(2002). The fact that the enzymes catalyzing the conversions of DHEA to androste-nedione, testosterone, and androst-5-ene-3β,17β-diol were not evidenced in the incubates with cultured preadipocytes, contrary to the experiment of Le Bail and collaborators (2002) on MCF-7 breast cancer cells, is probably due to a longer incubation time (20 h) in their case. A marked expression of these enzymes was not expected in the incubates with adipocytes, because 3β-HSD/Δ5-Δ4-isomerase and 17β-HSDs (types 2, 3, and 5), together with aromatase, appear to occur in adipose stromal cells and preadipocytes rather than in differentiated adipocytes (Belanger et al. 2002).
Although adipose tissue can uptake a variety of free and conjugated steroids, the most important precursor appears to be DHEA-S, which circulates at much higher concentrations than free DHEA, gonadal steroids, and corticoids (Labrie et al. 2001). We have found that the adipose tissue transcribes the genes encoding three OATPs, namely OATP-B, OATP-D, and OATP-E, but no other members of the OATP and OAT families. The same OATPs plus OAT-4 are also expressed in placental tissue, where OATP-B is the predominant mediator of Na+-dependent transport of DHEA-S (Ugele et al. 2003). The same transporters are also present in the human mammary gland (Pizzagalli et al. 2003) and temporal lobe (Steckelbroeck et al. 2004). Interestingly, we have measured similar rates of uptake and hydrolysis of DHEA-S by both preadipocytes and adipocytes, suggesting that fat cell differentiation is not associated with major changes in DHEA-S transport and desulfation. Thus, excessive adipose tissue accumulation may potentially extract substantial amounts of DHEA-S from the circulation to the detriment of other peripheral organs with a rather constant mass that are also dependent on this steroid, a situation that may be suspected when the adrenal production of DHEA-S is inherently low or declining with aging (Labrie et al. 2001).
While the expression of sulfotransferase genes was not observed in adipose tissue, STS expression at the mRNA, protein, and catalytic activity levels was instead conclusively established. Notably, fat cells do not utilize the main placental promoter leading to the partially translated first exon 1a, but display two differently spliced transcripts leading to two distinct untranslated first exons 0a and 0b, followed by a common partially translated exon 1b, placed at a greater distance from exon 2 than exon 1a. These two variants, in which the one with exon 0b seems to prevail over that with 0a, were also detected in the placenta as minor isoforms.
The use of different first exons combined with alternative promoters is a mechanism allowing spatial- and/or temporal-specific regulations of transcription, whose selectivity may be further enhanced by different sets of transcriptional regulatory factors in target cells. In many genes, like CYP19 encoding human cytochrome P450arom (Bulun et al. 2003), alternative splicing does not alter protein translation in different tissues. In the case of the STS gene, however, the location of the initiation codon in a different exon for placenta (1a) as compared to the isoforms 0a and 0b (1b) results in proteins with dissimilar N-terminal regions. Nevertheless, this should not interfere with the catalytic activity of the protein, because the N-terminal regions correspond to the signal peptide that is post-translationally cleaved during STS maturation (Stein et al. 1989).
A putative second initiation codon in frame with the remaining of the coding region was found in exon 2 of both placenta and adipose tissue cDNAs but, like the ATG in the placental exon 1a, its context matches only partially the Kozac consensus sequence, as it lacks a G base at position +4. If this alternative ATG is actually used, it would eliminate any sequence difference in the N-terminal region of placental and adipose STS proteins. A heterogeneity of human STS mRNA transcripts due to alternative polyadenylation sites has also been described (Ferrante et al. 2002), which may indirectly affect catalytic activity by influencing mRNA half-life.
The proximal promoter of placental STS gene is peculiar because it lacks a functional TATA box, as do the promoters of many housekeeping genes (Weis & Reinberg, 1992), but unlike them, it has a low GC content and lacks multiple binding sites for Sp1. It also seems to be not tightly regulated because it lacks binding sites for known transcription factors, though several distal regulatory elements have been characterized within the 1.3 kbp region upstream of the TSS (Li et al. 1996). A preliminary sequence analysis of the 5′-flanking region of each new first exon of adipose tissue failed to reveal a TATA box or an initiator element. Computer analysis of the proximal promoters with the TFSEARCH program (http://www.cbrc.jp/research/db/TFSEARCH.html) showed no consensus elements for transcription factors related to steroidogenesis, such as steroid receptors or StAR and SF-1. Notably, the program CpG islands (http://l25.itba.mi.cnr.it/cgi-bin/wwwcpg.pl) evidenced a 200 bp GC-rich region 50 nt upstream of exon 0b.
These data support the conclusion that the transcriptional regulation of adipose tissue STS differs from that of placental STS and disclose new avenues of investigation in the field of breast tumors, where STS expression is increased and may indirectly support breast tumor growth by boosting a local formation of estrogens.
Primer sequences used in RT-PCR and 5′- and 3′-RACE analyses
| Sequences | Positiona | GenBank no. | ||
|---|---|---|---|---|
| aNucleotide position in the reported sequence. | ||||
| bPrimer taken from Steckelbroeck et al.(2004). | ||||
| Primer | ||||
| SCC-1 | Sense | 5′-AAGACTTCACCCCATCTCCGTGAC-3′ | 1199→1222 | M14565 |
| SCC-2 | Antisense | 5′-ACCCCAGCCAAAGCCCAAGT-3′ | 1418→1399 | M14565 |
| C17-1 | Sense | 5′-CAATGAGAAGGAGTGGCACCA-3′ | 1263→1283 | NM_000102 |
| C17-2 | Antisense | 5′-CTTTGAAAGAGTCGATCAGAAAGAC–3′ | 1531→1507 | NM_000102 |
| AROM-1 | Sense | 5′-GAATCGGGCTATGTGGACGTGTTG-3′ | 590→613 | Y07508 |
| AROM-2 | Antisense | 5′-AGATGTCTGGTTTGATGAGGAGAG-3′ | 749→726 | Y07508 |
| SF1-1 | Sense | 5′-GGTGTCCGGCTACCACTACGG-3′ | 57→77 | U76388 |
| SF1-2 | Antisense | 5′-TGAAGCCATTGGCCCGAATCTG-3′ | 355→334 | U76388 |
| StAR-1 | Sense | 5′-CCAGGAGCTGGCCTATCTCCAG-3′ | 210→231 | BC010550 |
| StAR-2 | Antisense | 5′-CCAGGAGCTGGCCTATCTCCAG-3′ | 477→456 | BC010550 |
| E-SULT-1 | Sense | 5′-GAATGCAAAGGATGTGGCTGT-3′ | 493→513 | U08098 |
| E-SULT-2 | Antisense | 5′-TAATCCTGTCCACAAGCTCCTCT-3′ | 773→751 | U08098 |
| DHEA-SULT-1b | Sense | 5′-GTGGACAAAGCACAACTTCTG-3′ | 347→368 | NM_003167 |
| DHEA-SULT-2b | Antisense | 5′-TCTTACACAATGACCCCAGTC-3′ | 611→591 | NM_003167 |
| OATP-A-1 | Sense | 5′-GGAAGGACTAGAGACTAATGCTGA-3′ | 1026→1047 | NM_005075 |
| OATP-A-2 | Antisense | 5′-AGCATCAAGGAACAGTCAGGT-3′ | 1739→1719 | NM_005075 |
| OATP-B-1 | Sense | 5′-AGCCAGCCCAGACCCTCA-3′ | 274→291 | AB026256 |
| OATP-B-2 | Antisense | 5′-ACATGATCCCCACCACACTCA-3′ | 755→735 | AB026256 |
| OATP-C-1 | Sense | 5′-GATGGGTTGGAGCTTGGTG-3′ | 857→875 | AB026257 |
| OATP-C-2 | Antisense | 5′-CCTGCTAGACAGGGTGAGATGT-3′ | 1562→1541 | AB026257 |
| OATP-D-1 | Sense | 5′-GTGTGTGGGGCAGATGGCA-3′ | 1646→1664 | AB031050 |
| OATP-D-2 | Antisense | 5′-TTGTCTAGGGTCAGAGTAGAGGCA-3′ | 2223→2200 | AB031050 |
| OATP-E-1 | Sense | 5′-CGCTGCCTGCAGCTGCCA-3′ | 1606→1623 | AB031051 |
| OATP-E-2 | Antisense | 5′-AAGTTCCCGTGTGATTGCATCA-3′ | 2351→2330 | AB031051 |
| OATP-8-1 | Sense | 5′-GTTGGAGCTTGGTGGCTTGGT-3′ | 807→827 | NM_019844 |
| OATP-8-2 | Antisense | 5′-TACAGGGGATTGGTAAGGATGCT-3′ | 1057→1035 | NM_019844 |
| OAT-3-1 | Sense | 5′-TTCCCCATCTACATGGTCTTCC-3′ | 641→662 | AB042505 |
| OAT-3-2 | Antisense | 5′-TTACTTACGCCCATACCTGTTTGC-3′ | 1473→1450 | AB042505 |
| OAT-4-1 | Sense | 5′-GCAGCGTCTTCACCTCCA-3′ | 415→431 | AB026116 |
| OAT-4-2 | Antisense | 5′-GTCGAAGACCAGCCCATAGT-3′ | 1150→1131 | AB026116 |
| STS-1 | Sense | 5′-CTGACTTCTGTCACCACCCT-3′ | 637→656 | J04964 |
| STS-2 | Antisense | 5′-GTCCATGTTGCTAGTGGGCT-3′ | 1409→1390 | J04964 |
| STS-3 | Sense | 5′-CAACAGGATCACAAGCTGGA-3′ | 183→202 | J04964 |
| STS-4 | Antisense | 5′- AGGGTCAGGATTAGGGCTGCT -3′ | 889→869 | J04964 |
| STS-5 | Sense | 5′-ACCCTCATCTACTTCACAT-3′ | 1206→1224 | J04964 |
| STS-8 | Antisense | 5′- CACATGCGTCTGTCTGGT -3′ | 1985→1968 | J04964 |
| STS-6 | Antisense | 5′- AGGTGCTGAGTGAGTTTCACT -3′ | 409→389 | J04964 |
| STS-10 | Antisense | 5′- CAGTAGGAGGAAAGGGATCT -3′ | 242→223 | J04964 |
| STS-7 | Sense | 5′- AGGAAGCTGCGGACAGACACA -3′ | 1795→1815 | J04964 |
| STS-Pl/1a | Sense | 5′-CAGCTGTAGTGAGGTTGCA-3′ | 23→41 | J04964 |
| STS-Ad/0a | Sense | 5′-GAGAACCGCTACCATGCAG-3′ | 34→52 | AM072429 |
| STS-Ad/0b | Sense | 5′-GAAGAAGTCCGTCCATGTCA-3′ | 50→69 | AM072428 |
Target gene, PCR primers and conditions used in the RT-PCR analyses
| Primers | [Mg++] | Annealing (°C) | Extension time (s) | |
|---|---|---|---|---|
| Target gene | ||||
| P450scc | SCC-1-SCC-2 | 2 mM | 58 | 15 |
| P450c17 | C17-1–C17-2 | 2 mM | 56 | 20 |
| P450arom | AROM-1–AROM-2 | 2 mM | 58 | 15 |
| SF-1 | SF1-1–SF1-2 | 2 mM | 56 | 25 |
| StAR | StAR-1–StAR-2 | 2 mM | 56 | 20 |
| E-SULT | E-SULT-1–E-SULT-2 | 2 mM | 56 | 20 |
| DHEA-SULT | DHEA-SULT-1–DHEA-SULT-2 | 1.2 mM | 56 | 20 |
| OATP-A | OATP-A-1 OATP-A-2 | 1.5 mM | 54 | 40 |
| OATP-B | OATP-B-1 OATP-B-2 | 1.5 mM | 60 | 30 |
| OATP-C | OATP-C-1 OATP-C-2 | 1.5 mM | 54 | 40 |
| OATP-D | OATP-D-1 OATP-D-2 | 1.5 mM | 60 | 35 |
| OATP-E | OATP-E-1 OATP-E-2 | 1.5 mM | 60 | 40 |
| OATP-8 | OATP-8-1 OATP-8-2 | 1.5 mM | 54 | 20 |
| OAT-3 | OAT-3-1 OAT-3-2 | 1.5 mM | 60 | 40 |
| OAT-4 | OAT-4-1 OAT-4-2 | 1.5 mM | 60 | 40 |
| STS | STS-1–STS-2 | 1.5 mM | 56 | 40 |
| STS | STS-3–STS-4 | 1.5 mM | 60 | 40 |
| STS | STS-5–STS-8 | 1.5 mM | 60 | 40 |
| STS | STS-Pl/1a–STS10 | 2 mM | 56 | 20 |
| STS | STS-Ad/0a–STS10 | 2 mM | 56 | 15 |
| STS | STS-Ad/0b–STS10 | 2 mM | 60 | 20 |
Sulfatase activity measured by production of DHEA from DHEA-S substrate in adipose tissues and adipocytes
| Adipose tissue 1 (pmol/mg per hour) | Adipose tissue 2 (pmol/mg per hour) | Adipocytes (pmol/106 cells per hour) | Placentan (mol/mg per hour) | |||||
|---|---|---|---|---|---|---|---|---|
| Each experiment was performed in triplicate except the experiments marked with *, which were performed singly. | ||||||||
| Mean | s.d. | Mean | s.d. | Mean | s.d. | |||
| 1 mg×30 min | 71.67 | 5.77 | 175.00 | 57.66 | – | – | 7.05 | 0.25 |
| 1 mg×60 min | 75.83 | 6.29 | 141.67 | 18.08 | – | – | – | – |
| 1 mg×90 min | 60.56 | 6.75 | 117.22 | 17.95 | – | – | – | – |
| 2 mg×30 min | 87.50* | – | 143.75 | 19.44 | – | – | – | – |
| 2 mg×60 min | 106.25* | – | 121.90 | 10.85 | – | – | – | – |
| 0.5 mg×15 min | – | – | – | – | – | – | 6.59 | 0.89 |
| 0.5 mg×30 min | – | – | – | – | – | – | 7.85 | 1.02 |
| 0.5 mg×45 | – | – | – | – | – | – | 6.43 | 3.22 |
| 106 cells×60 min | – | – | – | – | 127.50* | – | – | – |
| 106 cells×120 min | – | – | – | – | 108.75* | – | – | – |
| 80.36 | 17.39 | 139.91 | 22.85 | 118.13 | 13.26 | 6.98 | 0.64 | |
Representative expression analyses by RT-PCR of cytochromes P450scc, P450c17, P450arom, StAR, SF-1, STS, E-SULT, and DHEA-SULT mRNAs in human adipose tissues (Ad1, Ad2, and Ad3), placenta (Pl), and adrenal cortex (AdCx), as described in Materials and Methods. MW, molecular weights; C−, no-template negative control (water).
Citation: Journal of Endocrinology 190, 1; 10.1677/joe.1.06811
Representative expression analyses by RT-PCR of the mRNAs of organic anion transporting polypeptides (OATP-A, -B, -C, -D, -E, OATP-8, OAT-3, and OAT-4) presumably involved in transmembrane transfer of DHEA-S in human adipose tissues (Ad1, Ad2, and Ad3), as described in Materials and Methods. MW, molecular weights; C−, no-template negative control (water).
Citation: Journal of Endocrinology 190, 1; 10.1677/joe.1.06811
Schematic diagram of the 5′-region variants of STS mRNAs primarily transcribed from alternative promoters in human placenta and adipose tissue. In the main placental transcript, the first ATG is located in the first exon 1a. Two transcripts are expressed in adipose tissue with two different untranslated first exons 0a and 0b and a common exon 1b bearing the first ATG. A second putative ATG is found in exon 2, common to all transcripts. Exons are depicted as boxes; introns are shown as horizontal lines.
Citation: Journal of Endocrinology 190, 1; 10.1677/joe.1.06811
Alignment of the 5′ regions found in adipose tissue (Ad/0a and Ad/0b with first exons 0a and 0b respectively) and placenta (Pl) STS transcripts. The putative ATG codons are boxed. Arrows show the intron positions.
Citation: Journal of Endocrinology 190, 1; 10.1677/joe.1.06811
RT-PCR analyses of STS mRNAs expressed in seven different human adipose tissues and one placenta with primer pairs specific for a common region in exon 2 and for different 5′-UTR variants (exons 1a, 0a, and 0b). MW, molecular weights; C−, no-template negative control (water).
Citation: Journal of Endocrinology 190, 1; 10.1677/joe.1.06811
Western immunoblot analyses of STS expression in placenta (Pl; 0.5 and 1 μg protein per lane) and six samples of adipose tissue (50 μg protein). Electrophoresis was performed on 10% polyacrylamide gels.
Citation: Journal of Endocrinology 190, 1; 10.1677/joe.1.06811
We are deeply grateful to Dr Bernhard Ugele (Frauenklinik-Innenstadt, University Hospital Munich, Munich, Germany), for the generous supply of the antiserum against human STS. Research was supported by Ricerca sanitaria finalizzata, grant no. 87/02 from the Veneto Region and 60% funding from the Ministry of the University and Scientific and Technological Research of Italy. The authors declare that there is no conflict of interest that would prejudice the impartiality of this scientific work.
References
Belanger C, Luu-The V, Dupont P & Tchernof A 2002 Adipose tissue intracrinology: potential importance of local androgen/estrogen metabolism in the regulation of adiposity. Hormone and Metabolic Research 34 737–745.
Bulun SE, Sebastian S, Takayama K, Suzuki T, Sasano H & Shozu M 2003 The human CYP19 (aromatase P450) gene: update on physiologic roles and genomic organization of promoters. Journal of Steroid Biochemistry and Molecular Biology 86 219–224.
Cha SH, Sekine T, Kusuhara H, Yu E, Kim JY, Kim DK, Sugiyama Y, Kanai Y & Endou H 2000 Molecular cloning and characterization of multispecific organic anion transporter 4 expressed in the placenta. Journal of Biological Chemistry 275 4507–4512.
Cha SH, Sekine T, Fukushima JI, Kanai Y, Kobayashi Y, Goya T & Endou H 2001 Identification and characterization of human organic anion transporter 3 expressing predominantly in the kidney. Molecular Pharmacology 59 1277–1286.
Clyne CD, Speed CJ, Zhou J & Simpson ER 2002 Liver receptor homologue-1 (LRH-1) regulates expression of aromatase in preadipocytes. Journal of Biological Chemistry 277 20591–20597.
Ferrante P, Messali S, Meroni G & Ballabio A 2002 Molecular and biochemical characterisation of a novel sulphatase gene: aryl sulphatase G (ARSG). European Journal of Human Genetics 10 813–818.
Frayn KN, Karpe F, Fielding BA, Macdonald IA & Coppack SV 2003 Integrative physiology of human adipose tissue. International Journal of Obesity 27 875–888.
Gibb W & Lavoie JC 1984 Kinetic studies on the formation of estrogens from dehydroepiandrosterone sulfate by human placental microsomes. Endocrinology 114 2323–2329.
Hagenbuch B & Meier PJ 2003 The superfamily of organic anion transporting polypeptides. Biochimica et Biophysica Acta (BBA)—Biomembranes 1609 1–18.
Hernandez-Guzman FG, Higashiyama T, Osawa Y & Ghosh D 2001 Purification, characterization and crystallization of human placental estrone/dehydroepiandrosterone sulfatase, a membrane-bound enzyme of the endoplasmic reticulum. Journal of Steroid Biochemistry and Molecular Biology 78 441–450.
Kershaw EE & Flier JS 2004 Adipose tissue as an endocrine organ. Journal of Clinical Endocrinology and Metabolism 89 2548–2556.
Kozak M 1996 Interpreting cDNA sequences: some insights from studies on translation. Mammalian Genome 7 563–574.
Labrie F, Luu-The V, Labrie C & Simard J 2001 DHEA and its transformation into androgens and estrogens in peripheral target tissues: intracrinology. Frontiers in Neuroendocrinology 22 185–212.
Le Bail JC, Lotfi H, Charles L, Pépin D & Habrioux G 2002 Conversion of dehydroepiandrosterone sulfate at physiological plasma concentration into estrogens in MCF-7 cells. Steroids 67 1057–1064.
Li XM, Alperin ES, Salido E, Gong Y, Yen P & Shapiro LJ 1996 Characterization of the promoter region of human steroid sulfatase: a gene which escapes X inactivation. Somatic Cell and Molecular Genetics 22 105–117.
Luu-The V, Dufort I, Paquet N, Reimnitz G & Labrie F 1995 Structural characterization and expression of the human dehydroepiandrosterone sulfotransferase gene. DNA and Cell Biology 14 511–518.
Martel C, Melner MH, Gagne D, Simard J & Labrie F 1994 Widespread tissue distribution of steroid sulfatase, 3β-hydroxysteroid dehydrogenase/Δ5-Δ4 isomerase (3β-HSD), 17β-HSD, 5α-reductase and aromatase activities in the rhesus monkey. Molecular and Cellular Endocrinology 104 103–111.
Miki Y, Nakata T, Suzuki T, Darnel AD, Moriya T, Kaneko C, Hidaka K, Shiotsu Y, Kusaka H & Sasano H 2002 Systemic distribution of steroid sulfatase and estrogen sulfotransferase in human adult and fetal tissues. Journal of Clinical Endocrinology and Metabolism 87 5760–5768.
Pizzagalli F, Varga Z, Huber RD, Folkers G, Meier PJ & St-Pierre MV 2003 Identification of steroid sulfate transport processes in the human mammary gland. Journal of Clinical Endocrinology and Metabolism 88 3902–3912.
Puche C, Jose M, Cabero A & Meseguer A 2002 Expression and enzymatic activity of the P450c17 gene in human adipose tissue. European Journal of Endocrinology 146 223–229.
Reed MJ, Purohit A, Woo LWL, Newman SP & Potter BVL 2005 Steroid sulfatase: molecular biology, regulation, and inhibition. Endocrine Reviews 26 171–202.
Steckelbroeck S, Nassen A, Ugele B, Ludwig M, Watzka M, Reissinger A, Clusmann H, Lutjohann D, Siekmann L, Klingmuller D & Hans VH 2004 Steroid sulfatase (STS) expression in the human temporal lobe: enzyme activity, mRNA expression and immunohistochemistry study. Journal of Neurochemistry 89 403–417.
Stein C, Hille A, Seidel J, Rijnbout S, Waheed A, Schmidt B, Geuze H & von Figura K 1989 Cloning and expression of human steroid-sulfatase, membrane topology, glycosylation, and subcellular distribution in BHK-21 cells. Journal of Biological Chemistry 264 13865–13872.
Strauss JF, III., Martinez F & Kiriakidou M 1996 Placental steroid hormone synthesis: unique features and unanswered questions. Biology of Reproduction 54 303–311.
Sugawara T, Holt JA, Driscoll D, Strauss JF, III., Lin D, Miller WL, Patterson D, Clancy KP, Hart IM, Clark BJ et al.1995 Human steroidogenic acute regulatory protein: functional activity in COS-1 cells, tissue-specific expression, and mapping of the structural gene to 8p11.2 and a pseudogene to chromosome 13. PNAS 92 4778–4782.
Suzuki T, Hirato K, Yanaihara T, Kadofuku T, Sato T, Hoshino M & Yanaihara N 1992 Purification and properties of steroid sulfatase from human placenta. Endocrinologia Japonica 39 93–101.
Thanaraj TA & Clark F 2001 Human GC-AG alternative intron isoforms with weak donor sites show enhanced consensus at acceptor exon positions. Nucleic Acids Research 29 2581–2593.
Ugele B, St-Pierre MV, Pihusch M, Bahn A & Hantschmann P 2003 Characterization and identification of steroid sulphate transporters of human placenta. American Journal of Physiology – Endocrinology and Metabolism 284 E390–E398.
Weis L & Reinberg D 1992 Transcription by RNA polymerase II: initiator-directed formation of transcription-competent complexes. FASEB Journal 6 3300–3309.
Yen PH, Marsh B, Allen E, Tsai SP, Ellison J, Connoly L, Neiswanger K & Shapiro LJ 1988 The human X-linked steroid sulfatase gene and a Y-encoded pseudogene: evidence for an inversion of the Y chromosome during primate evolution. Cell 55 1123–1135.
