Dependent mvdp/akr1-b7 Expression in Mouse Vas Deferens, But Is ...

0013-7227/02/$15.00/0 Printed in U.S.A.

Endocrinology 143(9):3435–3448 Copyright © 2002 by The Endocrine Society doi: 10.1210/en.2002-220293

A 77-Base Pair LINE-Like Sequence Elicits AndrogenDependent mvdp/akr1-b7 Expression in Mouse Vas Deferens, But Is Dispensable for Adrenal Expression in Rats* PIERRE VAL, ANTOINE MARTINEZ, ISABELLE SAHUT-BARNOLA, CLAUDE JEAN, ` RE, AND ANNE-MARIE LEFRANC GEORGES VEYSSIE ¸ OIS-MARTINEZ Unite´ Mixte de Recherche Centre National de la Recherche Scientifique, 6547 Physiologie Compareé et Endocrinologie Molećulaire, Universite´ Blaise Pascal Clermont II, Complexe Universitaire des Ce´zeaux, 63177 Aubiere, France Mvdp/akr1-b7 (mouse vas deferens protein/aldo-keto reductase 1-B7) encodes an enzyme responsible for detoxification of a steroidogenesis byproduct. MVDP/AKR1-B7 is expressed in both rat and mouse adrenal cortex under ACTH control, whereas strong androgen-dependent accumulation in the vas deferens is mouse specific. Comparison of the regulatory regions of the two orthologs reveals a strong identity, disrupted by acquisition of a 77-bp LINE-derived sequence in the mouse promoter. Although ACTH responsiveness is observed in both species, the absence of this 77-bp sequence in the rat is associated with changes in transcription initiation sites. Transfection studies demonstrate that the CCAAT/enhancer-binding protein and selective promoter factor 1-binding sites previously shown to be essential for cAMP/ACTH induction in

M

VDP/AKR1-B7 (mouse vas deferens protein/aldoketo reductase 1-B7), a member of the aldo-keto reductase superfamily, functions as a detoxification enzyme, reducing isocaproaldehyde produced by cholesterol sidechain cleavage in the first step of steroidogenesis (1). Reflecting its enzymatic function, the protein is expressed in mouse steroidogenic tissues, mostly in the zona fasciculata of the adrenal cortex (2) and to a lesser extent in the testis (Ref. 3 and this work) and the ovary (4). Recent reports of our group have shown that in adrenals, mvdp/akr1-b7 gene transcription is primarily controlled by the pituitary hormone ACTH through the cAMP pathway both in vivo and in cultured cells (3, 5). However, MVDP/AKR1-B7 was initially described in the mouse vas deferens, where its expression is controlled by testosterone at the transcriptional level (6, 7). In vivo and in vitro functional analyses of the mouse mvdp/ akr1-b7 regulatory regions indicate that both adrenal and vas deferens expression are driven by a single promoter, but require different trans-acting factors and partially overlapping cis-acting sequences for tissue-specific expression and Abbreviations: AR, Androgen receptor; ARE, androgen response element; C/EBP, CCAAT/enhancer-binding protein; DBD, DNA-binding domain; DHT, dihydrotestosterone; EMSA, electrophoretic mobility shift assay; GST, glutathione-S-transferase; hAR, human androgen receptor; Inr, initiator-like sequence; mvdp/akr1-b7, mouse vas deferens protein/aldo-keto reductase 1-B7; NF1, nuclear factor 1; pCMV, cytomegalovirus promoter; PKA, protein kinase A; RACE, rapid amplification of cDNA ends; RNase, ribonuclease; SF-1, steroidogenic factor-1; Sp1, selective promoter factor 1.

the mouse are consequently dispensable in the rat. Our data support the idea that the most striking change generated by this acquisition is the strong, androgen-dependent, vas deferens expression observed in mouse. 1) In rat vas deferens, rakr1-b7 expression is barely detectable and is not androgen sensitive. 2) Androgen receptor binds efficiently to an androgen response element within the 77-bp mouse-specific element. 3) Its insertion confers androgen sensitiveness to rakr1-b7 regulatory regions in an androgen response elementdependent manner in transient transfections. We propose that this acquired androgen-responsive region may be responsible for vas deferens androgen-regulated gene expression in vivo. (Endocrinology 143: 3435–3448, 2002)

hormonal control. In transgenic mice, a small promoter region (⫺510 to ⫹41) is sufficient to drive CAT expression in adrenals and confer control by ACTH (5, 8), whereas a –1804 to ⫹41 promoter fragment is necessary to reproduce all of the mvdp/akr1-b7 gene androgen-dependent features in vas deferens (5). In transfected Y1 adrenocortical cells, CCAAT/ enhancer-binding protein (C/EBP; ⫺61) and selective promoter factor 1 (Sp1; ⫺52) binding sites are implicated in the ACTH/cAMP responsiveness, whereas a cryptic binding site for the steroidogenic tissue-enriched factor steroidogenic factor-1 (SF-1; ⫺102) as well as a nuclear factor 1 (NF1; ⫺76) binding site are necessary for basal activity of the proximal promoter (⫺121/⫹41) (9). In steroid-responsive T47D mammary tumor cells, the ⫺121/⫹41 minimal promoter is sufficient to ensure androgen responsiveness via interaction of the androgen receptor (AR) with a proximal androgen response element (ARE) at position –111 (10, 11). Interestingly, mvdp/akr1-b7 expression in the adrenal gland is conserved among rodents (2, 3), whereas Western blot experiments have shown that the high androgen-dependent MVDP/ AKR1-B7 expression in the vas deferens is mouse specific (2, 12). Indeed, the protein is barely detected in rat, rabbit, and guinea pig vas deferens (2), although it is detectable at similar levels in all the inbred mice strains tested (CD-1; BALB/c; NMRI; C57; DBA; Snell dwarf; B6/D2 and B6/CBA) as well as in house mouse (5, 12). In this work we undertook a comparative study of mvdp/ akr1-b7 expression and regulatory regions between mouse

3435

3436

Endocrinology, September 2002, 143(9):3435–3448

and rat. The underlying idea was that such a study might unravel 1) conserved cis-acting sequences accounting for the conserved adrenal expression and 2) diverging sequences accounting for the androgen-dependent expression in mouse vas deferens. Our results show that although expression in the adrenals of both mouse and rat is controlled by ACTH, there are marked functional changes in the rat akr1-b7 proximal promoter. Despite high sequence identity, the two promoters differ by a short 77-bp region that is absent from the rat sequence. This is correlated to low testosterone-insensitive expression in the rat vas deferens. Using functional assays, we propose that androgen responsiveness of the akr1-b7 gene in mouse vas deferens originates from insertion of these 77 bp within its promoter. Materials and Methods Experimental animals Animal studies were conducted in agreement with the standards described by the NIH Guide for the Care and Use of Laboratory Animals as well as with the laws and regulations applicable to animal manipulations in France.

Antibodies and Western blot experiments For production of the L3 polyclonal antibody, rabbits were injected with a fusion of the last 17 amino acids of the MVDP/AKR1-B7 protein and the antibodies were obtained and tested as described previously (13). B263 monoclonal antibody was described previously (12). Both antibodies are specific for MVDP/AKR1-B7. Western blots were performed as previously described (9). Proteins were detected by either monoclonal antibody B263 (1:700) or L3 polyclonal antibody (1:3000).

PCR techniques Isolation of the coding region of rakr1-b7. Coding regions of rakr1-b7 cDNA were isolated by RT PCR, starting from 2 ␮g rat adrenal total RNAs, using Moloney murine leukemia virus reverse transcriptase (Promega Corp., Charbonnie`res, France) and rTaq polymerase (Takara BioWhittaker, Inc., Verviers, Belgium) according to the manufacturers’ instructions. 5⬘ Rapid amplification of cDNA ends (RACE) PCR and full-length rakr1-b7 cDNA isolation. 5⬘-Untranslated regions of rakr1-b7 transcripts were isolated by RACE as described previously (14), except for the points below. Ten micrograms of total RNAs from rat adrenals were used as the RT template. cDNAs were purified using the QiaQuick PCR purification kit (QIAGEN, Hilden, Germany). The oligonucleotidic single-stranded anchor was 5⬘-PO3-ACTATCGATTCTGGAACCTTCAGAGG-NH2–3⬘. The anchor PCR primer was 5⬘-CCACCTCTGAAGGTTCCAGAATCGATAG-3⬘. The rAKR1-B7-specific primers were 5⬘CTGATACACATAGGCACAGTCAAA-3⬘ (external primer, complementary to rakr1-b7 cDNA from 127–150, where A of ATG translation start codon is ⫹1) for the first 35 cycles and 5⬘-AATGGCTGCCTTCACAGCTTCCTT-3⬘ (internal primer, complementary to rakr1-b7 cDNA from 85–108) for the next 35 cycles. A first amplification round was carried out (35 cycles), followed by a second nested amplification round using internal primer (35 cycles). The PCR products were cloned into pGEM-T-Easy (Promega Corp.) for DNA sequencing. Three products differing by the lengths of their 5⬘ extremities were isolated by this procedure. An oligonucleotide named RT-Fwd (5⬘-AATAAAGCGTGCATTCTACACCAGG-3⬘) was derived from the longer transcript and used with RT-Rev (5⬘-CGGGATCCCGCATTTGCTCAGCCCCTCATTAGTG-3⬘) to amplify a full-length cDNA by RT PCR. RT-Rev is complementary to the 3⬘ extremity cloned in silico from EST databanks (GenBank accession no. AA925864). The PCR product was cloned into pGEM-T-Easy (Promega Corp.), producing pGEM-T-Easy-rakr1-b7 vector. Its sequence is deposited in GenBank (accession no. AF182168). Isolation of rat akr1-b7 regulatory regions: inverse PCR. 5⬘ Regulatory regions of rakr1-b7 were cloned by inverse PCR using oligonucleotides

Val et al. • Comparative Study of akr1-b7 Expression

complementary to the 5⬘-untranslated region isolated by 5⬘ RACE PCR. Genomic DNA was isolated from Wistar rat (Rattus norvegicus) liver and phenol/chloroform-purified as described previously (15). Ten micrograms of genomic DNA were digested by 20 U ScaI restriction endonuclease (cutting site unknown), and 50 ng digested DNA were self ligated in a 200-␮l final volume. One microliter of this reaction was used as a template for a first round of PCR amplification with primers IPL1 (5⬘-ATAGAGGAGGCGGACTCTGACCGC-3⬘) and IPR1 (5⬘-GACTTGCCAGCCTCCCACAGAG-3⬘), 2.5 U LA Taq polymerase (Takara), 8 ␮l of a 2.5 mm of each deoxy-NTP solution, and 2.5 mm MgCl2. Cycling parameters were 5 min at 94 C and 30 cycles composed of 94 C for 30 sec, 66 C for 1 min, and 72 C for 10 min. The samples were then incubated at 72 C for 10 min. One fiftieth of the reaction was used as a template for a second round of PCR using nested primers IPL2 (5⬘-AAGCAGCCTATCAGCACAGTGCGG-3⬘) and IPR2 (5⬘-CCTCCTCACAGAGGAAGCAGGC-3⬘) in the same conditions except for the annealing temperature (65 C instead of 66 C). This procedure allowed isolation of about 1400 bp. These were cloned in pGEM-T-Easy vector (Promega Corp.) and sequenced by Eurogentec Bel (Seraing, Belgium). About 760 bp corresponded to rakr1-b7 5⬘-flanking sequences (GenBank accession no. AF182372), whereas the remainder corresponded to the 5⬘ extremity of rakr1-b7 first intron (not deposited in databanks).

In situ hybridization In situ hybridization was performed as described previously (16). Sense and antisense probes were prepared according to the DIG RNA labeling kit protocol (Roche Molecular Biochemicals, Meylan, France). Sense or antisense rakr1-b7 probes are complementary to rakr1-b7 3⬘untranslated sequences from 951-1184 (ATG as ⫹1). P450 11␤ antisense probe is complementary to coding sequences from 544 – 888 (ATG as ⫹1).

Animals and hormonal treatments Ten adult male Wistar rats were castrated and killed 20 d later. Among them five were treated with chronically administered testosterone (150 ␮g/d; cyclodextrin-complexed testosterone, Sigma, StQuentin-Fallavier, France) through Alzet 2002 osmotic pumps (Charles Rivers, St-Aubin-les-Elbeuf, France) for 10 d before death. Five intact adult rats served as controls in these experiments. A second group of 16 adult Wistar rats was hypophysectomized (Iffa Credo, L’Arbresle, France) and killed 9 d later. They were treated in subgroups of 4 animals during 3 d before death, with either ACTH injection (Synacthe`ne Retard, Novartis Pharma, France; 13.5 U/d, im injection), testosterone injection (1 mg/d, sesame oil solution, sc injection), or ACTH and testosterone. Intact rats of the previous group served as controls. A third group of 8 animals was treated with chronically administered dexamethasone acetate (120 ␮g/d in polyethylene glycol through Alzet 2001 osmotic pumps) and killed 8 d later. Four of them were treated by daily injections of ACTH (Synacthe`ne Retard, 13.5 U/d, im injection) 3 d before death. Four intact animals served as controls. Control, castrated, and dexamethasone-treated animals were fed a standard diet, whereas the drinking water of hypophysectomized animals was supplemented with 10% glucose. Animals were killed by decapitation after slight anesthesia with Fluotane (5% in air). The tissues were immediately frozen in liquid nitrogen and stored at – 80 C until further experiments.

Northern blot analysis Total RNAs from adrenals of each animal were individually extracted with RNA Plus (Quantum Bioprobe, Montreuil-sous-Bois, France), and Northern blots were individually performed with 20 ␮g of the RNAs as described previously (9). Probes used were the full-length rakr1-b7 cDNA extracted from pGEM-T-Easy-rAKR1-B7 by a NcoI/PstI digestion or a mouse ␤-actin probe extracted from pGEM-7ZF-␤-actin by an EcoRI/ BamHI digestion. After rakr1-b7 hybridization, membranes were stripped and hybridized with ␤-actin to normalize the loading of RNAs. Most of the hybridization signals were quantified by phosphorimager analysis (Bio-Rad Laboratories, Inc.).


Ribonuclease (RNase) protection assay RNase protection assay was performed as described previously (9). The riboprobe used was complementary to rakr1-b7 sequences from ⫺78 to ⫹219, encompassing the putative transcription initiation sites. A dideoxy sequence (Amersham Pharmacia Biotech) of the probe template ran along the reactions provided a molecular size marker.

Genomic Southern blot Ten micrograms of genomic DNA extracted from Wistar rat livers were digested by at least 10 U of the following restriction endonucleases: BamHI, BclI, BglII, BstXI, EcoRI, PvuII, SacI, ScaI, SphI, HindIII, and XbaI. Resulting fragments were resolved on a 0.7% agarose gel and blotted onto a nylon membrane. Genomic DNA for rakr1-b7 was detected by a probe complementary to the first 399 nucleotides of the gene’s first intron isolated by inverse PCR. Hybridization signals were revealed by a 4-d exposure to an x-ray film with two intensifying screens at – 80 C.

Electrophoretic mobility shift assay (EMSA) EMSA was performed as described previously (9). Recombinant glutathione-S-transferase (GST)-AR DNA-binding domain (DBD) was as described previously (17). Anti-Sp1 and anti-C/EBP␤ are commercial antisera (Santa Cruz Biotechnology, Inc., Santa Cruz, CA); anti-CTF1/ NF1 was provided by Dr. Naoko Tanese, New York University Medical Center (New York, NY). Oligonucleotides used in this study were: rat NF1, 5⬘-tgaggctgcttgccagtgcg-3⬘ (⫹64, ⫹83); mouse NF1, 5⬘-tgtggctgcttgccaatgtg-3⬘ (⫺78, ⫺59); rat C/EBP, 5⬘-ttgccagtgcggtcagagtcc-3⬘ (⫹72, ⫹93); mouse C/EBP, 5⬘-ttgccaatgtggtaagagccc-3⬘ (⫺69, ⫺48); rat Sp1, 5⬘-ggtcagagtccgcctcctctatcc-3⬘ (⫹83, ⫹106); mouse Sp1, 5⬘-ggtaagagcccgcctcctttatcc-3⬘ (⫺59, ⫺36); ⫺111 ARE, 5⬘-agcttgacatgaagttcctgttctcatggtcga-3⬘ (⫺117, ⫺93); ⫺111 AREm, 5⬘-agcttgacatgaagttcctttctcatggtcga3⬘; and C3(1) ARE, 5⬘-agcttacatagtacgtgatgttctcaagg-3⬘.

Luciferase reporter constructs A 754-bp TATA⫹ fragment (⫺584, ⫹170) encompassing the putative TATA box was amplified by PCR with 650 ng Wistar rat genomic DNA as a template. Primers were: forward, 5⬘-CGGGGTACCCCGAGTACATGGTATTTCATAATCATA-3⬘; and TATA⫹, 5⬘-GAAGATCTTCTGCAAAGTATGCAGATGAAATG-3⬘. A 688-bp TATA⫺ fragment (⫺584, ⫹104) devoid of the TATA box and its downstream sequences was amplified by PCR with 650 ng Wistar rat genomic DNA as a template. Primers were forward and TATA⫺ (5⬘-GAAGATCTTCCCTGGATAGAGGAGGCGG-3⬘). A 639-bp initiator-like sequence-positive (Inr⫹) fragment (⫺584, ⫹60) devoid of the putative NF-1, C/EBP, and Sp1 binding sites was amplified by PCR with 50 ng of the TATA⫹ plasmid as a template. Primers were forward and Inr⫹ (5⬘-agcacagtgcggcctccc3⬘). A 552-bp Inr⫺ fragment (⫺584, ⫺28) devoid of the three putative Inr sequences was amplified by PCR with 50 ng of the TATA⫹ plasmid as a template. Primers were forward and Inr⫺ (5⬘-ggtcatccaagatgagttggtaagt-3⬘). The resulting PCR products were cloned into pGL3 basic vector (Promega Corp.). A 830-bp TATA⫹::77-bp construct was obtained by inverse PCR using Herculase DNA polymerase (Stratagene, Amsterdam, The Netherlands) starting with 15 ng of the TATA⫹ plasmid as a template. Each half of the mvdp/akr1-b7 77-bp sequence was added to the 5⬘ end of each of the PCR primers ⫹ARE forward (5⬘-tgacatgaagttcctgttctcatgccccaacccttggc tgaggctgcttgccagtgcg-3⬘) and ⫹ARE reverse (5⬘atcttgtgccacactgatcat cgtaaacaaagattaaagctcagcacagtgcggcctcc-3⬘). This led to insertion of the mouse-specific 77 bp between nucleotides ⫹63 and ⫹64 of the rakr1-b7 regulatory regions (Fig. 2). The PCR product was gel purified and phosphorylated. Twenty nanograms of the linear product were then self-ligated overnight and transformed into DH5␣ bacterial strain. The 830-bp TATA⫹::77-bpm construct was obtained by mutating the ARE contained within the mouse-specific 77-bp sequence using the Gene Editor kit (Promega Corp.). The mutant oligonucleotide used was 5⬘-agcttgacatgaagttcctttctcatggtcga-3⬘. (The underline shows the mutated nucleotide.) 0.5 Luc construct was obtained by subcloning the 0.5 kb mvdp/akr1-b7 promoter (10) into the SmaI and BglII restriction sites of the pGL3 vector. 0.5 Luc AREm was obtained by mutating the ARE by the method described above. All constructs were sequenced to confirm

Endocrinology, September 2002, 143(9):3435–3448 3437

the correct orientation and nucleotide sequence of the akr1-b7 cloned regulatory regions.

Cell transfection and luciferase assay Y1 and T47D cells were maintained in DMEM-Ham’s F-12 medium (1:1) supplemented with 10% fetal calf serum, 2 mm l-glutamine, 100 U/ml penicillin, and 100 ␮g/ml streptomycin. Y1 and T47D cells were transfected at a density of 300,000 cells/well in 6-well plates (Falcon 35-3502, Merck Eurolab, Strasbourg, France). Y1 cells were transfected 24 h after seeding by the calcium phosphate method as described previously (9) with 1 ␮g of the luciferase reporter constructs (pGL3 TATA⫹, pGL3 TATA⫺, pGL3 Inr⫹, pGL3 Inr⫺)/well. Twelve hours after transfection, the effect of the cAMP pathway on the reporter constructs expression was assayed by addition of 10⫺5 m forskolin (in dimethylsulfoxide) for 24 h to the culture medium. T47D cells were transfected in serum-free medium, 24 h after seeding, with Exgen 500 (Euromedex, Mundolsheim, France), following the manufacturer’s instructions. One microgram of the reporter constructs (pGL3 TATA⫹, pGL3 TATA⫹::77 bp, pGL3 TATA⫹::77 bp, 0.5 Luc, 0.5 Luc ⫹ AREm) were transfected alone or cotransfected with 1 ng cytomegalovirus promoter (pCMV) human AR [expressing human AR (hAR)] or 1 ng pCMX (empty vector) using 4 ␮l Exgen 500/␮g DNA. Twelve hours after transfection, the effect of androgens on the reporter expression was assayed by addition of 10⫺6 m dihydrotestosterone (DHT; in ethanol) to the serum-free medium for 24 h. Luciferase assays were performed as described previously (9). Luciferase activity was corrected by the activity of a cotransfected pCMV ␤-galactosidase plasmid. Each experiment was repeated at least 3 times (12 points). Results were expressed as the mean of 12 points ⫾ sd.

Results AKR1-B7 protein tissue distribution

Recent data have shown that MVDP/AKR1-B7 is a major reductase for isocaproaldehyde generated by the conversion of cholesterol to pregnenolone in mouse adrenals (1). We examined the expression of the protein in rat and mouse steroidogenic tissues using a specific polyclonal antibody. Vas deferens was included as a control (Fig. 1). A specific signal at 34.5 kDa (the size of MVDP/AKR1-B7) was detected in adrenals, vas deferens, testis, and ovary in both rat and mouse, although differences were observed in the levels of protein accumulation. Indeed, expression in the adrenals and testes is comparable between the two species, whereas expression in the vas deferens, which is the major site of expression in the mouse, is barely detectable in the rat. Isolation of the cDNA encoding rAKR1-B7

A protein corresponding to rat AKR1-B7 is thus easily detected by a specific antibody in rat adrenals. We undertook the isolation of the corresponding cDNA by RT-PCR and 5⬘ RACE PCR, starting from total rat adrenals RNA samples. A

FIG. 1. AKR1-B7 protein expression in different rat and mouse tissues. Sixty micrograms of proteins were extracted from different rat and mouse tissues, electrophoresed in a 10% SDS-PAGE gel, transferred to a nitrocellulose membrane, and incubated with L3 antibody (1:3000) specific for mouse MVDP/AKR1-B7.

3438


full-length cDNA (1431 bp long), referred to hereafter as rakr1-b7, has been isolated (GenBank accession no. AF182168). The open reading frame is 948 nucleotides long, and an eukaryotic consensus polyadenylation signal (AATAAA) is observed at position 1416, suggesting that the 3⬘-untranslated region of the isolated cDNA is complete. The first in-frame methionine codon is contained within the translation initiation consensus sequence, ACCATGA. This assignment predicts a translation product of 316 amino acids, the same size as MVDP/AKR1-B7 (18). Comparison of the amino acid sequences showed that rAKR1-B7 has 90.8% overall identity with MVDP/AKR1-B7, and that the conservation of residues is uniformly spread along the protein. The 5⬘- and 3⬘-untranslated regions of rakr1-b7 cDNA showed 50% and 67.5% identities with those of mvdp/akr1-b7, respectively. Residues shown to be important for aldo-keto reductase function are perfectly conserved between MVDP/ AKR1-B7 and rAKR1-B7 (1), and enzymatic properties are equivalent (data not shown). It can be concluded that rAKR1-B7 is encoded by the rat ortholog of mvdp/akr1-b7. Isolation and characterization of the 5⬘-flanking regions of the rakr1-b7 gene

Upstream rakr1-b7 flanking regulatory regions were cloned by an inverse PCR strategy using primers complementary to the 5⬘-untranslated region obtained by 5⬘ RACE PCR. About 800 bp of the 5⬘-flanking region were sequenced (GenBank accession no. AF182372) and provided the basis for analysis of the rakr1-b7 gene promoter. A computerassisted alignment of the sequences of the 5⬘-flanking DNA of rat and mouse genes has been established (Fig. 2). The first striking observation is that the alignment between the two sequences is interrupted over 77 bp, between nucleotides –154 and –78 of the mouse sequence. Despite this major difference, the two promoters display an overall extensive sequence similarity. Indeed, they share 77.5% and 84.5% identity upstream and downstream, respectively, of the alignment interruption. When comparing the mvdp/akr1-b7 and rakr1-b7 sequences by aligning the ATG codons, classical promoter features, such as a TATA box (CATAAA) and a CAAT box, are conserved in both species. However, three initiator-like sequences (Inr: consensus sequence, 5⬘PyPya⫹1Nt/aPyPy-3⬘) (19) are present at positions –24, ⫹10, and ⫹35 only in the rat gene (Fig. 2). On the basis of sequence comparisons, binding sites for Sp1 (⫺52), CCAAT/ enhancer-binding protein, C/EBP (⫺61), NF1 (⫺76), activating protein 1 (⫺373), and steroidogenic factor-1, SF-1 (⫺503 to ⫺489, ⫺458 and ⫺249) are conserved in both species. Interestingly, the mvdp/akr1-b7-proximal ARE sequence from position –111 to –97, a cryptic SF-1-binding site at position ⫺102, and a NF1 half-binding site at position – 82, lie within the 77-bp region that is absent from the rat upstream sequences. Rakr1-b7 is specifically expressed in the zona fasciculata of the adrenal cortex

In the mouse, MVDP/AKR1-B7 has been shown to be restricted to the zona fasciculata of the adrenal cortex by immunohistochemical (2) and in situ hybridization experi-


ments (8). The tissular localization of the rakr1-b7 transcripts on rat adrenal sections was determined by in situ hybridization using digoxigenin-labeled riboprobes (Fig. 3). As a control, cytochrome P-450 11␤-hydroxylase, which catalyzes the transformation of deoxycorticosterone to corticosterone, was included as a zone-specific enzyme, expressed in the zona fasciculata-reticularis and not in the zona glomerulosa (20). Expression of rakr1-b7 mRNA was restricted to the zona fasciculata of the rat adrenal cortex. No signal was observed with rakr1-b7 sense-labeled probe, confirming the specificity of the detection. Hormonal regulation of rakr1-b7 mRNA accumulation in rat adrenals

MVDP/AKR1-B7 expression in mouse adrenals is under the control of androgens (5) and ACTH (2). The effects of both hormones on the rat mRNA accumulation was assayed by Northern blot experiments. Twenty micrograms of total RNAs from adult rat adrenals were subjected to Northern blots and hybridized with a full-length rakr1-b7 cDNA probe (Fig. 4A). A single mRNA species was detected with a size of 1.4 kb that was similar to that previously reported for mouse adrenocortical cells (2) and mouse adrenals (9). Quantitation of the Northern blot signals (Fig. 4B) showed that a 9-d hypophysectomy markedly reduced rakr1-b7 mRNA accumulation (2.5-fold). Administration of ACTH for 3 d, to males hypophysectomized 6 d previously, was sufficient to restore control levels of rakr1-b7 mRNA. Similarly, the amounts of rakr1-b7 mRNA were strongly reduced after dexamethasone treatment (5 d) and were restored by ACTH administration (3 d). The effect of the androgen status of the animal on rakr1-b7 transcript accumulation was then assayed. Neither castration nor testosterone injection to castrated or hypophysectomized animals had any significant effect on mRNA accumulation (Fig. 4B). Furthermore, testosterone had no effect on ACTH responsiveness. Collectively, these results demonstrate that ACTH is the main regulator of rakr1-b7 gene expression in rat adrenals. Analysis of the transcription initiation sites of the rakr1-b7 gene in the adrenals

Preliminary 5⬘ RACE PCR experiments that were conducted to isolate a full rakr1-b7 transcript, allowed the cloning of three 5⬘ ends (data not shown). To determine the transcription initiation site(s) from the rat promoter, we performed RNase protection experiments on total RNAs from rat adrenals. Three major protected fragments were found surrounding the putative Inr (initiator) elements. Their intensities were similar, suggesting that the corresponding transcription start sites were used with equivalent efficiency. A much weaker band was found downstream of the putative TATA box (Fig. 5A, Ctrl lane). To rule out the possibility that the different transcription initiation sites might be the product of different rakr1-b7 genes, rat DNA was analyzed by Southern blot hybridization with a probe complementary to rakr1-b7. In the different digestions, one labeled band of high intensity was observed (Fig. 5B), indicating that the rakr1-b7 gene is a single copy gene. To determine which of the transcription start sites were affected by ACTH regulation,



FIG. 2. Sequence comparison between mvdp/akr1-b7 and rakr1-b7 5⬘-flanking regions. Rat akr1-b7 5⬘-flanking sequences obtained by inverse PCR (bold italic) are aligned with mouse mvdp/akr1-b7 5⬘-flanking sequences (Roman). Nucleic acid identity is shown by colons; the absence of alignment is shown by asterisks. Vertical arrows show the extent of the probe used for mapping of 5⬘ ends of the transcripts. Vertical triangles stand for the transcription initiation sites detected by RNase mapping experiments. The shaded gray triangle represents a minor transcription initiation site. The ⫹1 position in the rat sequence was arbitrarily set at the more distal transcription initiation site. Transcription initiation in mouse is shown by a thin horizontal arrow and ⫹1 numbering. Putative Inr (PyPyA⫹1NT/APyPy) are indicated by large horizontal open arrows. Brackets show the extent of the constructs tested in Fig. 6. Horizontal arrows show the extent of the probes used in the EMSA experiments in Fig. 6. Regulatory protein-binding sites identified in the flanking regions are boxed. AP 1, Activating protein 1.

RNase protection experiments were performed with RNAs extracted from adrenals of animals that were treated by dexamethasone or dexamethasone plus ACTH (same animals as in Fig. 4). The intensity of the three major protected frag-

ments was strongly decreased by dexamethasone administration and enhanced in dexamethasone-treated animals that were supplemented with ACTH. The intensity of the band that mapped downstream the putative TATA box was not

3440


FIG. 3. rakr1-b7 is expressed in the zona fasciculata of the adult male rat adrenals. P450 11␤ and rakr1-b7 transcripts were detected in situ on sections of adult male rat adrenals with the corresponding antisense digoxigenin-labeled riboprobes. rakr1-b7 probe is complementary to the 3⬘-untranslated region. Sense rakr1-b7 probe hybridization is shown as a negative control. ZF, Zona fasciculata; ZG, zona glomerulosa; ZR, zona reticularis.

modulated by ACTH levels. We therefore conclude that in adrenals, rakr1-b7 transcription is initiated and hormonally regulated at three predominant sites clustered in an 80-bp region containing the three putative initiators. An adenosine residue mapping to the more distal initiation site was numbered ⫹1 (refer to Fig. 2). NF1, C/EBP, and Sp1 binding sites of the rakr1-b7 5⬘-flanking regions are able to bind their respective transcription factors

In the mouse adrenal gland, mvdp/akr1-b7 expression is under the control of ACTH, which activates steroidogenesis through the cAMP/protein kinase A (PKA) pathway (2, 5). In mouse Y1 adrenocortical cells, MVDP/AKR1-B7 expression has been shown to be up-regulated by increases in intracellular cAMP levels (2). Binding sites for C/EBP (⫺61) and Sp1 (⫺52) are essential for this induction, whereas cryptic SF-1 (⫺102) and NF1 (⫺76) binding sites are required for basal expression (9). On the basis of sequence comparisons (Fig. 2), NF1 (⫹66), C/EBP (⫹81), and Sp1 (⫹90) binding sites are conserved in the rat promoter, but are located downstream of the transcription initiation. This raises the question of their functionality in rakr1-b7 expression. Their ability to bind their respective transcription factors was thus analyzed by EMSA. Incubation of the 32P-labeled rat-NF1 probe with nuclear extracts from Y1 cells, produced two comigrating retarded complexes that were displaced by a 50-fold molar excess of the cold probe itself, by a 50-fold molar excess of the mouse NF1 probe, or partly by the addition of an antibody raised against CTF-1/NF1 (Fig. 6A). This indicates that the rat NF1 site is indeed able to bind NF1 proteins contained within Y1 nuclear extracts. The ability of the C/EBP and Sp1 sites to bind their cognate transcription factors was demonstrated in the same way (Fig. 6A).


FIG. 4. Effect of hormonal treatments on rakr1-b7 accumulation in rat adrenals. A, Northern blot experiments. A first group of adult male rats (⫻10) was castrated for 20 d (Castrated) or castrated for 10 d and treated with testosterone for the next 10 d (Castrated ⫹ T) before death. A second group (x16) was hypophysectomized for 9 d (Hpx) or hypophysectomized for 6 d and treated with ACTH (Hpx ⫹ ACTH) and/or testosterone (Hpx ⫹ ACTH ⫹T/Hpx ⫹ T) for the next 3 d before death. The third group of animals (x8) was treated with dexamethasone acetate for 8 d (Dex) or treated with dexamethasone for 8 d and ACTH for 3 d before death (Dex ⫹ ACTH). Twenty micrograms of total RNAs extracted from each animal adrenals were then subjected to Northern blots and hybridized with a probe specific to rakr1-b7 cDNA. Blots were stripped and reprobed with a ␤-actin probe to normalize RNA loading. A typical experiment is shown. B, Quantitation of Northern blot signals. Individual hybridization signals were quantitated by phosphorimager analysis. The histograms represent the mean values of rakr1-b7 hybridization signals standardized with ␤-actin and expressed relative to untreated animals signals ⫾ SD.

Different cis elements are required for transcriptional activation of mvdp/akr1-b7 and rakr1-b7 in adrenocortical cells

To test the ability of rakr1-b7 sequences to direct basal and hormone-stimulated expression, constructs containing 5⬘-flanking sequences from the rakr1-b7 gene (⫺584, ⫹170) or 3⬘ to 5⬘ deletions of these sequences linked to the luciferase reporter gene in a promoterless vector were transfected into murine Y1 adrenocortical cells that naturally express MVDP/AKR1-B7. As shown in Fig. 6B, pGL3 TATA⫹ construct harboring the whole regulatory sequences isolated by inverse PCR conferred a strong basal level of expression to the luciferase gene in Y1 adrenocortical cells. This indicates that the rat akr1-b7 flanking regions constitute a functional promoter in adrenocortical cells. Deletion of the sequences from ⫹105 to ⫹170, including the putative TATA box, resulted in a slight decrease in the basal level of expression (pGL3 TATA⫺). Further deletion of the sequences from ⫹61 to ⫹104 encompassing binding sites for NF1, C/EBP, and Sp1 (pGL3 Inr⫹) resulted in a further decrease in the basal expression of the reporter, indicating that all of these regions enclose cis-acting sequences involved in the basal expression of the



FIG. 5. rakr1-b7 transcription is predominantly initiated at Inr sequences. A, Transcription initiation mapping. Adult male rats were untreated (Ctrl), treated with dexamethasone acetate (Dex), or treated with dexamethasone acetate and ACTH (Dex ⫹ ACTH; same animals as in Fig. 4), and total RNAs were extracted from adrenals. Three micrograms were then hybridized with a 32P-labeled antisense riboprobe complementary to rakr1-b7 upstream regions. Single-stranded RNAs were then digested by 3 U RNase. Digestion resistant hybrids were resolved on a sequencing gel. The molecular size marker is provided by a dideoxy sequence of the probe (lanes GATC, in inverse complementary orientation). Specific complexes are indicated by arrows. Arrowhead shows the excess undigested probe. B, rakr1-b7 is a single copy gene. Ten micrograms of rat genomic DNA were digested with at least 10 U of the restriction endonucleases named above. Digested fragments were electrophoresed, blotted, and hybridized with a probe complementary to rakr1-b7. The membrane was autoradiographed for 4 d at – 80 C.

gene. Finally, deletion of the Inr regions from ⫺27 to ⫹60 reduced expression of the reporter gene (pGL3 Inr⫺) to a level comparable to that of empty vector (pGL3 basic). To study a possible regulation of rakr1-b7 promoter activity by the PKA pathway, transfected Y1 cells were cultured 24 h in the absence or presence of forskolin, an activator of cAMP synthesis. Addition of forskolin to the medium resulted in a significant 2.5-fold increase in luciferase activity for the longer construct (pGL3 TATA⫹), indicating that the rakr1-b7 regulatory regions are responsive to cAMP. Deletion of the TATA box and downstream regions (pGL3 TATA⫺) resulted in a marked increase in the forskolin inducibility of the construct (4.4- vs. 2.5-fold induction), which was further increased by deletion of the regions encompassing the NF1-, C/EBP-, and Sp1-binding sites (pGL3 Inr⫹, 6.0 vs. 4.4). Deletion of the Inr regions (pGL3 Inr⫺) abrogated the induced expression of the reporter. The stimulatory effect of forskolin was not observed with the control pGL3 basic vector. Collectively, these data indicate that cAMP stimulation of the rakr1-b7 gene is achieved essentially at the transcriptional level.

Although the ⫹105 to ⫹170 and ⫹61 to ⫹104 DNA regions are required for maximal basal expression of rakr1-b7 promoter in Y1 cells, these are dispensable for cAMP/ACTH induction. Thus, the elements required for the transcriptional activation of the akr1-b7 gene differ between rat and mouse; the TATA box at position ⫹111 is not essential for the expression of rakr1-b7, whereas C/EBP- and Sp1-binding sites are not required for induction of the expression by cAMP. Hormonal regulation of rAKR1-B7 protein expression in vas deferens

Castration experiments have shown that MVDP/ AKR1-B7 expression is under the strict control of androgens in the mouse vas deferens (5, 7, 21). To test whether the very low expression of rakr1-b7 gene in rat vas deferens was influenced by androgens, proteins of the vas deferens from adult males that were untreated, castrated, (20 d), or castrated and supplemented with testosterone (20 d castration, testosterone the last 10 d) were submitted to West-

3442



FIG. 6. Control of akr1-b7 expression by the cAMP pathway requires distinct sequences in mouse and rat proximal promoters. A, NF1 (⫹66), C/EBP (⫹81), and Sp1 (⫹90) binding sites contained in the rat akr1-b7 flanking regions bind their respective factors. Six micrograms of nuclear extracts from Y1 cells (Y1) were incubated with 90,000 cpm of a 32P-labeled double-stranded oligonucleotide corresponding to the NF1, C/EBP, and Sp1 binding sites of the rat promoter in the absence or presence of a 50-fold molar excess of unlabeled competitor oligonucleotides. Competitors were either the binding sites themselves (name preceded by r) or the corresponding previously characterized binding sites of the mouse promoter (name preceded by m) (9). Antibodies directed against CTF1/NF1 (␣NF1), C/EBP␤ (␣C/EBP) or Sp1 (␣Sp1) were added as indicated. Asterisks indicate supershifted complexes. Arrows indicate the specific complexes; arrowheads show the free probe. N.S., Not specific. B, Transfection of rakr1-b7 5⬘-flanking regions in the Y1 adrenocortical cell line. The fulllength rakr1-b7 isolated promoter (pGL3 TATA⫹); the promoter devoid of the TATA box (pGL3 TATA⫺); the promoter devoid of the NF1-, C/EBP-, and Sp1-binding sites (pGL3 Inr⫹); or the promoter devoid of the Inr sequences (pGL3 Inr⫺) were cloned upstream of the luciferase reporter gene. One microgram of each of the constructs was transfected into Y1 cells. Effect of 10⫺5 M forskolin treatment on reporter expression was assayed for 24 h (⫹Fsk, dark gray bars). Fold induction by forskolin is indicated above the bars. Empty reporter vector was included as a control (pGL3 basic). Luciferase activity was corrected by the ␤-galactosidase activity of a cotransfected pCMV ␤-galactosidase plasmid. Values are expressed as the mean of at least three experiments (12 points) ⫾ SD. Different letters stand for significantly different values (P ⬍ 0.005) determined by t test.

ern blot analysis. As depicted in Fig. 7A, the amounts of rAKR1-B7 protein in rat vas deferens were not affected by these treatments, indicating that the low expression of rAKR1-B7 in the rat vas deferens is not under androgenic control. Similar results were also obtained at the level of rakr1-b7 mRNA (data not shown).

Androgen sensitivity of akr1-b7 expression is dependent on the presence of a 77-bp mouse-specific androgen-responsive region

Sequence comparisons show that a 77-bp region containing an ARE (⫺111) and a half-binding site for NF1 (⫺82),



FIG. 7. The androgen-insensitive rakr1-b7 gene can be imposed androgen sensitiveness by insertion of a 77-bp androgen-responsive sequence present in mvdp/akr1-b7 promoter. A, Expression of rAKR1-B7 is not controlled by androgens in the rat vas deferens. Adult male rats were treated as indicated below. Proteins from vas deferens (30 ␮g) were electrophoresed in a 10% SDS-PAGE gel, transferred to a nitrocellulose membrane, and incubated with B263 monoclonal antibody (1:700) specific for mouse MVDP/AKR1-B7. Castrated, 20 d castration; T, testosterone; Ctrl, control untreated animal. B, Recombinant AR binds mvdp/akr1-b7 proximal ARE (⫺111) in EMSA. Three picomoles of recombinant GST-AR DBD were incubated with 90,000 cpm of a 32Plabeled double-stranded oligonucleotide corresponding to the mvdp/akr1-b7 ⫺111 ARE or a mutant version in the absence or presence of a 30-fold molar excess of unlabeled competitor oligonucleotides. Competitors were either the binding sites themselves or the previously characterized ARE of the C3(1) gene. Arrows indicate the specific complexes; the asterisk stands for the supershift; arrowheads show the free probe. C, Insertion of a mouse 77-bp androgen-responsive sequence confers androgen sensitivity to rakr1-b7 5⬘regulatory regions. The –584/⫹170 wildtype rakr1-b7 promoter (pGL3TATA⫹), the same sequences with insertion of the mouse-specific mvdp/akr1-b7 77-bpm sequence between nucleotides ⫹63 and ⫹64 (pGL3 TATA⫹::77 bp), and the same construct with mutation of the ARE (pGL3 TATA⫹::77 bpm) were cloned upstream of the luciferase reporter gene. One microgram of each construct was transfected into T47D cells. Constructs harboring the –510/⫹41 promoter of the mouse mvdp/ akr1-b7 gene with wild-type (0.5 Luc) or mutated ARE (0.5 Luc AREm) were included as controls. One nanogram of pCMV hAR expressing the hAR or 1 ng of the corresponding pCMX empty vector were cotransfected as indicated. Effect of a 10⫺6 M DHT treatment on reporter expression was assayed for 24 h (⫹DHT, dark gray bars). Fold induction by DHT treatment is indicated above the histograms. Luciferase activity was corrected by the ␤-galactosidase activity of a cotransfected pCMV ␤-galactosidase plasmid. Values are expressed as the mean of at least 3 experiments (12 points) ⫾ SD.

which are essential for androgenic induction in vitro (10, 17), is absent from the otherwise highly conserved rat 5⬘-flanking regions (Fig. 2). Because mvdp/akr1-b7 expression in the mouse vas deferens is strictly dependent on the presence of androgens, it was tempting to correlate the low unregulated expression of the gene in rat vas deferens to the absence of the 77-bp mouse-specific sequence in the rakr1-b7 5⬘-flanking regions. The ability of the –111 ARE contained within these 77 bp to bind AR was assayed by EMSA. As shown in Fig.

7B, recombinant GST-AR DBD efficiently binds the –111 ARE. The binding is specific, as evidenced by displacement of the retarded complex either by the ARE itself and the C3(1) ARE (22) or by the anti-GST antibody. A mutant version of the –111 ARE was unable to bind AR DBD or to compete for binding on the wild-type probe. Thus, the ARE contained within the 77-bp mouse-specific sequence is able to bind AR. We then assayed the ability of the 77-bp mouse-specific sequence to confer androgenic sensitivity to the rat regula-

3444


tory regions. The pGL3 TATA⫹ construct, containing the wild-type rakr1-b7 5⬘-flanking regions from –584 to ⫹170 driving the expression of the luciferase reporter gene, was transfected in the T47D mammary carcinoma cell line (Fig. 7C). Basal expression of the construct was not affected by cotransfection of the hAR (pCMV hAR) or the corresponding empty vector (pCMX). Treatment of the cells with 10⫺6 m DHT for 24 h in the presence or absence of hAR did not modify the level of expression of the construct compared with that in unstimulated cells. This demonstrates the absence of androgen sensitivity of rakr1-b7 wild-type regulatory regions. As a control, the expression of the luciferase reporter gene driven by the 0.5-kb promoter (⫺510 to ⫹41) of the mouse mvdp/akr1-b7 gene (0.5 Luc) was induced 10.6fold after treatment with 10⫺6 m DHT in the presence of hAR. As previously described (10), this induction was abrogated by mutation of the ARE at –111 (0.5 Luc AREm). The mousespecific 77-bp sequence was then introduced between nucleotides ⫹63 and ⫹ 64 of the –584/⫹170 rakr1-b7 promoter construct (pGL3 TATA⫹::77 bp). This construct displayed a basal expression equivalent to that of the wild-type construct, which was unaffected by cotransfection of pCMX or pCMVhAR. DHT treatment caused a 5.1- to 6.4-fold induction of the construct activity in the absence of exogenous AR, probably due to endogenous AR expression in T47D cells (23). A 14.1-fold increase was observed in the presence of cotransfected hAR, demonstrating that the 77 bp themselves were able to confer androgen sensitivity to the rakr1-b7 regulatory regions. To confirm the essential role of AR and its cognate ARE in this androgen sensitivity, the ARE was mutated in the rakr1-b7 construct (pGL3 TATA⫹::77 bpm). Basal expression of the construct was not affected by the mutation. However, the DHT responsiveness of this mutant construct was nearly abolished (2.2- vs. 14.1-fold) in the presence or absence of exogenous AR. Thus, the 77-bp sequence has the capacity to transfer androgen responsiveness in an AREdependent manner. These data suggest that absence of the 77-bp mouse-specific sequence could account for the absence of androgen responsiveness of the rat akr1-b7 gene in vivo and consequently to the low vas deferens expression in this species. Reciprocally, the 77-bp mouse-specific sequence comprising the proximal ARE could be responsible for androgenic sensitivity of the mouse mvdp/akr1-b7 gene. Origin of the mouse-specific 77-bp androgenresponsive sequence

The above results demonstrate the need of a 77-bp androgen-responsive sequence in mouse akr1-b7 promoter for androgen-dependent vas deferens expression. In other mammals, including guinea pig, rabbit, and rat, AKR1-B7 protein was undetectable or was present at very low levels in the vas deferens (2). Although we only isolated the mouse and rat akr1-b7 promoters, this suggests an active insertion of the 77-bp sequence in the 5⬘-flanking regions of the mouse mvdp/akr1-b7 gene rather than a deletion in the five other species. An extensive search of the 77-bp androgen-responsive sequence in the genomic databases using a FASTA algorithm (24) returned one significant 71% identity hit with a genomic region of human chro-


mosome 22 within the putative regulatory region of a septin-3-like gene (accession no. HS250D10/Z99716; Fig. 8A). Interestingly, this portion of the chromosome was annotated as a LINE-like region. We thus undertook a comparison between our 77-bp sequence and the repetitive elements sequences contained within RepBase (25). The 77-bp sequence matched a L1_MC4 sequence within the rodent repetitive sequences database (release 3.4.2) with a 65% identity score, indicating that indeed, LINE sequences were contained within the 77-bp androgen-responsive sequence (Fig. 8B). However, the relatively low conservation as well as the lack of typical LINE transposition signatures such as a polyadenylation signal and direct repeats flanking the insertion site, indicate that the 77-bp sequence is not the direct product of a simple transposition event (26). Rather, one can speculate that this LINE-like sequence was probably brought to the mvdp/akr1-b7 promoter by a secondary recombination event after the rat/mouse speciation. Discussion Species-specific mechanisms are involved in ACTH regulation of akr1-b7 gene

The present work has focused on the identification of elements involved in hormonal and tissue-specific regulation of the mvdp/akr1-b7 gene through a comparative study between mouse and rat. The expression of MVDP/AKR1-B7 is known to be under the control of ACTH in adrenals (2, 5) and gonadotropins in other steroidogenic tissues (3, 4). These hormones acting via increases in cAMP and subsequent activation of the PKA signaling pathway stimulate steroidogenesis acutely by mobilizing cholesterol to the inner mitochondrial membrane and chronically by inducing the expression of steroidogenic genes (27). Here we have shown that rakr1-b7 gene expression in rat adrenals is also under the control of ACTH. Transfection experiments in murine Y1 adrenocortical cells show that the 5⬘-flanking sequences of the rakr1-b7 gene are able to confer cAMP responsiveness to a reporter gene, indicating that stimulation of the rakr1-b7 gene by the PKA pathway is, at least in part, achieved at the transcriptional level. In mouse vas deferens and adrenals, transcription is initiated approximately 30 bp downstream of a TATA box (9, 28). In the rat, although the TATA box is conserved, transcription is mainly initiated at three points located around three Inrs that display a marked sensitivity to ACTH induction. As expected, deletion of the Inr regions completely abolishes the expression of the rat promoter constructs, whereas deletion of the TATA box and downstream sequences does not preclude cAMP induction and only slightly affects their basal expression, thus confirming the essential role of the putative Inrs in rakr1-b7 expression. Previous transfection experiments have indicated that the Sp1and C/EBP␤-binding sites, located at positions –52 and – 61, respectively, are required for cAMP trans-activation of the mvdp/akr1-b7 proximal promoter (⫺120/⫹41) in Y1 cells (9). The change in the transcription initiation site in the rat promoter places the conserved binding sites for C/EBP (⫹81) and Sp1 (⫹90) downstream from the transcription start site. Despite their conserved ability to interact with their cognate



FIG. 8. The 77-bp mouse-specific sequence is part of a degenerated LINE sequence. A, FASTA search results. The 77-bp sequence was submitted to a FASTA search against the nonredundant DNA database at GenBank. The search returned a significant 71% identity hit with BAC sequence HS250D10/Z99716, upstream of a septin-3-like gene on human chromosome 22. The sequence alignment is shown. The ARE and half-NF1 binding sites are boxed. The thin horizontal arrows indicate the 5⬘ to 3⬘ sequence orientation. A schematic shows the location of the alignment within HS250D10. The thick horizontal arrow is the mouse-specific 77-bp sequence. The cDNA for septin-3-like sequence is shown as a large gray rectangle. Numbering is Z99716 numbering as defined in GenBank. 124914 is the putative ATG initiation codon. B, FASTA alignment of the 77-bp mouse-specific sequence with the rodent repetitive DNA sequence database (RepBase). The sequence search returned a significant 65% identity with the L1_MC4 DNA consensus, corresponding to a subclass of Mus musculus LINE L1 3⬘ end. At the top, the schematic depicts the relative positions of the L1, L1_MC4 (thin horizontal arrows), and 77-bp sequence (thick horizontal arrow). The arrows show the 5⬘ to 3⬘ sequence orientation. Numbering is as defined in RepBase. The sequence alignment is shown below. The ARE and half-NF1 binding sites are boxed. N, Undefined nucleotide; Y, pyrimidine.

transcription factors, their deletion enhances cAMP induction of the reporter expression, providing evidence that the proximal binding sites for C/EBP and Sp1 are clearly dispensable for induction of rakr1-b7 expression by the cAMP pathway. It has been reported that SF-1 mediates basal and cAMPinduced expression of many steroidogenic genes, including the cholesterol side-chain cleavage cytochrome P450 gene (CYP11A), the aromatase cytochrome P450 gene (CYP19), the 17␣-hydroxylase/C17–20 lyase gene (CYP17), the human 3␤hydroxysteroid dehydrogenase ⌬5 34 isomerase, and the steroidogenic acute regulatory protein (29 –37). Interestingly, in vivo ACTH responsiveness of mvdp/akr1-b7 in adrenals is conferred by sequences from –510 to ⫹41 of the mvdp/akr1-b7 gene (5, 8), harboring four putative SF-1-binding sites (⫺503, ⫺458, ⫺249, and ⫺102). Although a cryptic SF-1-binding site

(⫺102) contained within the mouse-specific 77-bp sequence is absent from the rat promoter, the remaining sites could account for the conserved level of adrenal expression despite the functional differences in the proximal promoters in rat and mouse. Their role in the cAMP-induced adrenal expression of akr1-b7 is currently being investigated. Role of a 77-bp androgen-responsive sequence in androgendependent vas deferens expression of the mouse mvdp/akr1-b7 gene

An important observation of this study is the barely detectable constitutive expression of the rakr1-b7 gene in the rat vas deferens, whereas very high levels of expression and absolute androgen requirement are distinctive features of the mouse gene in this tissue. Previous transient transfections

3446


have demonstrated that sequences from –121 to ⫹41 contribute strongly to basal and androgen-dependent mvdp/ akr1-b7 gene transcription (10, 11). Studies using deoxyribonuclease I footprinting, band-shift assays, and site-directed mutagenesis have shown the presence of a functional ARE at positions –111 to –97 that confers androgen responsiveness to the heterologous TK promoter (this paper and Refs. 10, 17, and 38). It was thus tempting to correlate the almost absence of vas deferens expression and the complete lack of androgenic control in rat to the lack of a 77-bp mouse-specific sequence harboring the functional ARE (⫺111) in otherwise highly conserved regulatory regions. Indeed, the mousespecific 77-bp sequence is able to confer androgen sensitivity to the insensitive rat regulatory regions, on an ARE-dependent basis, in transfection assays. As expression in the mouse vas deferens is strictly dependent on the presence of androgens (3, 5, 7), the absence of the mouse-specific 77-bp sequence appears to account for the absence of androgen responsiveness of the rat akr1-b7 gene in vivo and consequently to its very low vas deferens expression. A new model of acquired androgen regulation

DNA sequences responsible for regulation by androgens have been found in many inducible genes by DNA transfection techniques. However, the results of such experiments do not necessarily reflect the physiological cellular regulation because of limitations in the cotransfection assays and the absence of an appropriate chromatin structure of the constructs. Transgenic studies have been used to characterize regulatory sequences involved in the tissue-specific and androgen-regulated expression of genes in the kidney (39 – 41), prostate (42, 43), epididymis (44), vas deferens, and adrenals (5). Although large regulatory regions containing information required for hormonal regulation have been delineated in these studies, there is no direct evidence that AREs play a role as elements of androgen action in vivo. Valuable information on the role of AR and the cognate AREs in a physiological context was obtained from model genes in which androgen dependence was acquired during evolution. The Slp (sex-limited protein) is the most documented model of this kind; acquisition of a 120-bp androgen-specific enhancer contained within the 5⬘ long terminal repeat of a retrovirus has imposed a sex-specific expression in the kidney and liver of male mice (45). However, androgenic regulation of Slp is a complex process, as androgens act directly on Slp expression only in the kidney (46, 47). Furthermore, the 120-bp enhancer is a composite sequence with one major ARE, termed HRE3, and numerous other trans-acting sequences that contribute either positively or negatively to androgen induction through the recruitment of AR and nonreceptor ubiquitous proteins (48, 49). At last, although some congenic mouse haplotypes display either no expression or constitutive expression of Slp, these still carry the retrovirus insertion, indicating that the retrovirus alone might not be sufficient for the sex-specific expression in vivo (45, 50). Another model of the acquisition of androgen responsiveness is the Gus-s gene. Different mouse haplotypes coexist, ranging from no response of Gus-s to androgens (Gusor) to high response to androgens (Gusa) (51). Interestingly, the androgen-


sensitive phenotype correlates to the presence in the ninth intron of the Gus-s gene of a DNase I-hypersensitive site associated with a consensus HRE that binds an as yet unidentified kidney-specific factor. In the nonresponsive mice this region is absent (52). However, as indicated by transgenic mice experiments, Gus-s ninth intron only protects androgen responsiveness mediated by the 3.8-kb promoter and is not by itself essential for the hormonal responsiveness of the transgene in vivo (39). The last model that may provide interesting information is the kap gene, whose expression is regulated by androgens in mouse kidney (40, 53). Recent experiments have shown the presence of a large insertion of a L1 LINE sequence, required for expression of kap promoter transgenes, in the regulatory regions of the gene. However, the role of this sequence in androgen responsiveness has not been elucidated (54). In all these model genes the sequences responsible for the androgenic sensitivity have been either broadly delineated or are quite complex enhancers. By contrast, our model provides evidence that the acquisition of androgen responsiveness resulted from the insertion of a very simple, but functionally active, 77-bp androgen-responsive region containing a single ARE. A selective advantage for high mvdp/akr1-b7 expression in mouse vas deferens?

What could be the selective advantage for the high mvdp/ akr1-b7 expression in the vas deferens accounting for the maintenance of the 77-bp androgen-responsive region? Isocaproaldehyde, the major substrate for MVDP/AKR1-B7, is not produced in the vas deferens. However, AKR1-B7 is also able to undertake the detoxification of the lipid peroxidation product 4-hydroxynonenal (1, 3). This substrate is produced in the vas deferens and needs to be eliminated to prevent its toxic effects on DNA and proteins (55). Thus, in mouse vas deferens, 4-hydroxynonenal might be degraded by AKR1B7, whereas in species where MVDP expression is low, this detoxification might be undertaken by other members of the aldose reductase or alcohol dehydrogenase families. Alternatively, AKR1-B7 activity may have been recruited for an as yet unidentified function in mouse vas deferens. Interestingly, the production of prostaglandins E and F is elevated in mouse vas deferens compared with that in other rodents (56, 57), and it has been reported that rat aldose reductase could be involved in the synthesis of prostaglandin E2 by the kidney (58). As MVDP/AKR1-B7 shares 70% amino acid identity with mouse aldose reductase, it is tempting to speculate that it could take part in the high vas deferens prostaglandin production in the mouse. In summary, the present data provide new insights into the acquisition of hormonal responsiveness during evolution. They provide functional evidence that the presence of a mouse-specific 77-bp androgen-responsive region containing an ARE may be implicated in androgen-induced transcription of a target gene in vivo. Acknowledgments The authors thank Drs C. White, P. Lachaume, and P. Vernet for critical reading of this manuscript, and J. P. Saru for excellent technical assistance.


Received March 11, 2002. Accepted May 28, 2002. Address all correspondence and requests for reprints to: Dr. Antoine Martinez, Unite´ Mixte de Recherche Centre National de la Recherche Scientifique, 6547 Physiologie Compare´ e et Endocrinologie Mole´ culaire, Universite´ Blaise Pascal, Clermont II, Complexe Universitaire des Ce´ zeaux, 24 avenue des Landais, 63177 Aubie`re Cedex, France. E-mail: [email protected]. * The rakr1-b7 promoter and cDNA sequences are deposited in GenBank under accession nos. AF182372 and AF182168, respectively.


21.

22.

23.

References

24.

1. Lefrançois-Martinez A. M., Tournaire C, Martinez A, Berger M, Daoudal S, Tritsch D, Veyssie`re G, Jean Cl 1999 Product of side-chain cleavage of cholesterol, isocaproaldehyde, is an endogenous specific substrate of mouse vas deferens protein, an aldose reductase-like protein in adrenocortical cells. J Biol Chem 274:32875–32880 2. Aigueperse C, Martinez A, Lefrançois-Martinez AM, Veyssie`re G, Jean Cl 1999 Cyclic AMP regulates expression of the gene coding for a mouse vas deferens protein related to the aldo-keto reductase superfamily in human and murine adrenocortical cells. J Endocrinol 160:147–154 3. Martinez A, Aigueperse C, Val P, Dussault MH, Tournaire C, Berger M, Veyssie`re G, Jean Cl, Lefrançois-Martinez AM 2001 Physiological functions and hormonal regulation of mouse vas deferens protein (AKR1/B7) in steroidogenic tissues. Chem Biol Interact 130 –132:903–917 4. Brockstedt E, Peters-Kottig M, Hegele-Hartung C, Lessl M 2000 Luteinizing hormone induces mouse vas deferens protein expression in the murine ovary. Endocrinology 141:2574 –2581 5. Martinez A, Lefrançois-Martinez AM, Manin M, Guyot S, Jean-Faucher Ch, Veyssie`re G, Kahn A, Jean Cl 1999 5⬘-flanking and intragenic sequences confer androgenic and developmental regulation of mouse aldose reductase-like gene in vas deferens and adrenal in transgenic mice. Endocrinology 140:1338 –1348 6. Martinez A, Berger M, Jean-Faucher Ch, Veyssie`re G, Jean Cl 1993 Radioimmunoassay of mouse vas deferens protein: developmental profile and hormonal regulation. J Endocrinol 138:99 –105 7. Martinez A, Berger M, Jean Cl 1990 Androgen regulation of the mRNA encoding a major protein of the mouse vas deferens. Mol Cell Endocrinol 72:201–211 8. Sahut-Barnola I, Lefrançois-Martinez AM, Jean C, Veyssie`re G, Martinez A 2000 Adrenal tumorigenesis targeted by the corticotropin-regulated promoter of the aldo-keto reductase AKR1B7 gene in transgenic mice. Endocr Res 26: 885– 898 9. Aigueperse C, Val P, Pacot C, Darne C, Lalli E, Sassone-Corsi P, Veyssie`re G, Jean Cl, Martinez A 2001 SF-1 (steroidogenic factor-1), C/EBP ␤ and ubiquitous transcription factors NF1 and SP1 are required for regulation of the mouse aldose reductase-like gene (AKR1/B7) expression in adrenocortical cells. Mol Endocrinol 15:93–111 10. Fabre S, Manin M, Pailhoux E, Veyssie`re G, Jean Cl 1994 Identification of a functional androgen response element in the promoter of the gene for the androgen-regulated aldose reductase-like protein specific to the mouse vas deferens. J Biol Chem 269:5857–5864 11. Fabre S, Darne Ch, Veyssie`re G, Jean Cl 1995 Characterization of the promoter of the gene for a mouse vas deferens protein related to the aldo-keto reductase superfamily: effect of steroid hormones and phorbol esters. J Steroid Biochem Mol Biol 55:315–325 12. Taragnat C, Berger M, Jean Cl 1990 Tissue and species specificity of mouse ductus deferens protein. J Androl 11:279 –286 13. Morel L, Brochard D, Manin M, Simon AM, Jean C, Veyssie`re G 2001 Mouse seminal vesicle secretory protein of 99 amino acids (MSVSP99): characterization and hormonal developmental regulation. J Androl 22:549 –557 14. Apte AN, Siebert PD 1993 Anchor-ligated cDNA libraries: a technique for generating a cDNA library for the immediate cloning of the 5⬘ ends of mRNAs. Biotechniques 15:890 – 893 15. Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K 1999 Preparation of genomic DNA from mammalian tissue. In: Ausubel FM, ed. Short protocols in molecular biology, 4th Ed. New York: Wiley & Sons; 2-9 –2-10 16. Braissant C, Whali W 1998 A simplified in situ hybridization protocol using non-radioactively labelled probes to detect abundant and rare mRNAs on tissue sections. Biochemica 1:10 –16 17. Darne Ch, Morel L, Claessens F, Manin M, Fabre S, Veyssie`re G, Rombauts W, Jean Cl 1997 Ubiquitous transcription factors NF1 and SP1 are involved in the androgen activation of the mouse vas deferens protein promoter. Mol Cell Endocrinol 132:13–23 18. Pailhoux EA, Martinez A, Veyssie`re GM, Jean CG 1990 Androgen-dependent protein from mouse vas deferens. cDNA cloning and protein homology with the aldo-keto reductase superfamily. J Biol Chem 245:19932–19936 19. Javahery R, Khachi A, Lo K, Zenzie-Gregory N, Smale ST 1994 DNA sequence requirements for transcriptional initiator activity in mammalian cells. Mol Cell Biol 14:116 –127 20. Ogishima T, Suzuki H, Hata JI, Mitani F, Ishimura Y 1992 Zone-specific

25. 26. 27. 28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38. 39.

40.

41.

42.

43.

44.

expression of aldosterone synthase cytochrome P-450 and cytochrome P-450 11␤ in rat adrenal cortex: histochemical basis for the functional zonation. Endocrinology 130:2971–2977 Taragnat C, Berger M, Jean Cl 1988 Preliminary characterization, androgendependence and ontogeny of an abundant protein from mouse vas deferens. J Reprod Fertil 83:835– 842 Celis L, Claessens F, Peeters B, Heyns W, Verhoeven G, Rombauts W 1993 Proteins interacting with an androgen-responsive unit in the C3(1) gene intron. Mol Cell Endocrinol 94:165–172 Takane KK, McPhaul MJ 1996 Functional analysis of the human androgen receptor promoter. Mol Cell Endocrinol 119:83–93 Pearson W, Lipman D 1988 Improved tools for biological sequence comparison. Proc Natl Acad Sci USA 85:2444 –2448 Jurka J 1998 Repeats in genomic DNA: mining and meaning. Curr Opin Struct Biol 8:333–337 Smit A 1999 Interspersed repeats an other mementos of transposable elements in mammalian genomes. Curr Opin Genet Dev 9:657– 663 Simpson ER, Waterman ER 1988 Regulation of the synthesis of steroidogenic enzymes in adrenocortical cells by ACTH. Annu Rev Physiol 50:427– 440 Pailhoux E, Veyssie`re G, Fabre S, Tournaire C, Jean Cl 1992 The genomic organization and DNA sequence of the mouse vas deferens androgen-regulated protein gene. J Steroid Biochem Mol Biol 42:561–568 Liu Z, Simpson ER 1997 Steroidogenic factor 1 (SF-1) and Sp1 are required for regulation of bovine CYP11A gene expression in bovine luteal cells and adrenal Y1 cells. Mol Endocrinol 11:127–137 Bakke M, Lund J 1995 Mutually exclusive interactions of two nuclear orphan receptors determine activity of a cyclic adenosine 3⬘, 5⬘-monophosphate-responsive sequence in the bovine CYP17 gene. Mol Endocrinol 9:327–339 Carlone DL, Richards JS 1997 Functional interactions, phosphorylation and levels of 3⬘, 5⬘-cyclic adenosine monophosphate-regulatory element binding protein and steroidogenic factor-1 mediate hormone-regulated and constitutive expression of aromatase in gonadal cells. Mol Endocrinol 11:292–304 Zhang P, Mellon SH 1997 Multiple orphan nuclear receptors converge to regulate rat P450c17 gene transcription: novel mechanisms for orphan nuclear receptor action. Mol Endocrinol 11:891–904 Wooton-Kee CR, Clark BJ 2000 Steroidogenic factor-1 influences proteindeoxyribonucleic acid interactions within the cyclic adenosine 3⬘,5⬘-monophosphate-responsive regions of the murine steroidogenic acute regulatory protein gene. Endocrinology 141:1345–1355 Morohashi KI, Zanger UM, Honda SI, Hara M, Waterman MR, Omura T 1993 Activation of CYP11A and CYP11B gene promoters by the steroidogenic cellspecific transcription factor, Ad4BP. Mol Endocrinol 7:1196 –1204 Zhang P, Mellon SH 1996 The orphan nuclear receptor steroidogenic factor 1 regulates the cyclic adenosine 3⬘,5⬘-cyclic-monophosphate-mediated transcriptional activation of rat cytochrome P450c17 (17␣-hydrolase/C17–20 lyase). Mol Endocrinol 10:147–158 Mason JL, Keeny DS, Bird IM, Rainey WE, Morohashi KI, Leers-Sucheta S, Melner MH 1997 The regulation of 3␤-hydroxysteroid dehydrogenase expression. Steroids 62:164 –168 Clark BJ, Soo SC, Caron KM, Ikeda Y, Parker KL, Stocco DM 1995 Hormonal and developmental regulation of the steroidogenic acute regulatory protein. Mol Endocrinol 9:1346 –1355 Darne Ch, Veyssie`re G, Jean Cl 1998 Phorbol ester causes ligand-independent activation of the androgen receptor. Eur J Biochem 256:541–549 Ding Y, Sigmund C 2001 Androgen-dependent regulation of human angiotensinogen expression in KAP-hAGT transgenic mice. Am J Physiol Renal Physiol 280:F54 –F60 Ding Y, Davisson RL, Hardy DO, Zhu LJ, Merrill DC, Catterall JF, Sigmund CD 1997 The kidney androgen-regulated protein promoter confers renal proximal tubule cell-specific and highly androgen-responsive expression on the human angiotensinogen gene in transgenic mice. J Biol Chem 272:28142–28148 Thornton S, Thomas DW, Gallagher PM, Ganschow RE 1998 Androgen responsiveness of mouse kidney ␤-glucuronidase requires 5⬘-flanking and intragenic Gus-s-sequences. Mol Endocrinol 12:333–341 Greenberg NM, Demayo FJ, Sheppard PC, Barrios R, Lebovitz R, Finegold M, Angelopoulo R, Dodd JG, Duckworth ML, Rosen JM, Matusik RJ 1994 The rat probasin gene promoter directs hormonally and developmentally regulated expression of a heterologous gene specifically to the prostate in transgenic mice. Mol Endocrinol 8:230 –239 Cleutjens KBJM, Vander Korput HAGM, Ehren-Van Eekelen CC, Sikes RA, Fasciana C, Chung LW, Trapman J 1997 A 6Kb promoter fragment mimics in transgenic mice the prostate-specific and androgen-regulated expression of the endogenous prostate-specific antigen gene in humans. Mol Endocrinol 11: 1256 –1265 Lareyre JJ, Thomas TZ, Zheng WL, Kasper S, Ong DE, Orgebin-Crist MC, Matusik RJ 1999 A 5-kilobase pair promoter fragment of the murine epididymal retinoic acid-binding protein gene drives the tissue-specific, cell-specific, and androgen-regulated expression of a foreign gene in the epididymis of transgenic mice. J Biol Chem 274:8282– 8290

3448


45. Stavenhagen JB, Robins DM 1988 An ancient provirus has imposed androgen regulation on the adjacent mouse sex-limited protein gene. Cell 55:247–254 46. Nelson S, Robins D 1997 Two distinct mechanisms elicit androgen-dependent expression of the mouse sex-limited protein gene. Mol Endocrinol 11:460 – 469 47. Georgatsou E, Bougarel P, Meo T 1993 Male-specific expression of mouse sex-limited protein requires growth hormone, not testosterone. Proc Natl Acad Sci USA 90:3626 –3630 48. Adler A, Danielsen M, Robins D 1992 Androgen-specific gene activation via a consensus glucocorticoid response element is determined by interaction with non receptor factors. Proc Natl Acad Sci USA 89:11660 –11663 49. Adler A, Scheller A, Hoffman Y, Robins D 1991 Multiple components of a complex androgen-dependent enhancer. Mol Endocrinol 5:1587–1596 50. Nelson S, Robins D 1997 Regulatory capacity of an androgen-specific enhancer of the mouse Slp gene in transgenic mice. Mol Cell Endocrinol 133:89 –97 51. Lund S, Miller D, Chapman V, Ganschow R 1988 Androgen regulation of murine ␤-glucuronidase expression: identification and characterization of a non-response variant. Genetics 119:151–156 52. Lund S, Gallagher P, Wang B, Porter S, Ganschow R 1991 Androgen re-


53.

54. 55.

56.

57. 58.

sponsiveness of the murine ␤-glucuronidase gene is associated with nuclease hypersensitivity, protein binding and haplotype-specific sequence diversity within intron 9. Mol Cell Biol 11:5426 –5433 Meseguer A, Catterall J 1990 Cell-specific expression of kidney androgenregulated protein messenger RNA is under multihormonal control. Mol Endocrinol 4:1240 –1248 Hardy D, Niu E, Catterall J 2001 Kap promoter analysis in vivo: a regulatory role for a truncated L1 repeat. Mol Cell Endocrinol 181:57– 67 Comporti M 1998 Lipid peroxidation and biogenic aldehydes: from the identification of 4-hydroxynonenal to further achievments in biopathology. Free Radic Res 28:663– 635 Marshburn P, Clark M, Shabanowitz R 1989 Immunohistochemical localization of prostaglandin H synthase in the epididymis and vas deferens of the mouse. Biol Reprod 41:491– 497 Badr F 1974 Prostaglandin levels in tissues of the male reproductive system in six strains of mice. Endocrinology 96:540 –543 Zager P, Dorin R, Shah V, Kaplan D, Mann P, Frey H, Kellner T 1993 Aldose reductase expression and prostaglandin E2 production are coordinately regulated in cultured rat mesengial cells. Metabolism 42:269 –276