| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Environmental and Molecular Fish Biology Group, Department of Biological Sciences, Hatherly Laboratories, University of Exeter, Exeter, Devon EX4 4PS, United Kingdom
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and ERß), and in some species the Esr2 subtype has two forms, Esr2b (formerly ERß1) and Esr2a (formerly ERß2 or ER
). There is little information, however, on the different characteristics and functional significance of the two receptor subtypes in fish, and this is especially relevant for understanding the disruption of ER signaling by chemicals with estrogenic activity. In this study, the full-length cDNAs for esr1 (3167 base pairs [bp]) and esr2b (2318 bp), and a partial-length (267 bp) cDNA for esr2a, were cloned and characterized in fathead minnow (fhm; Pimephales promelas), and their patterns of expression established during development and in adults. Real-time polymerase chain reaction revealed some clear distinctions in the ontogenic and tissue expression of fhm esr1, esr2b, and esr2a, suggesting different functions for each ER subtype. Fhm ERs were expressed in brain, pituitary, liver, gonad, intestine, and gill of male and female fish, esr2b and esr2a were also expressed in muscle. Fhm esr1 and esr2b were expressed predominantly in the liver, whereas fhm esr2a was expressed predominantly in intestine and was lowest expressed in liver. Responses of the different hepatic ERs in male fathead minnow exposed to 100 ng estradiol/L differed, with a significant induction (5-fold) of fhm esr1 but no effect on esr2b or esr2a expression, suggesting different mechanisms of regulation for the different ERs. The detailed characterization of ERs in fathead minnow provides the foundation for understanding the molecular basis of estrogenic disruption in fish.
developmental expression, endocrine disruption, estrogen receptors, fathead minnow, real-time polymerase chain reaction, tissue expression
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
ERs possess the hallmark modular structure characteristic of the steroid/thyroid hormone receptor superfamily [8, 9] and, on the basis of amino acid sequence homology, are divided into six functional domains designated A–F [10]. The C domain (DNA binding-domain; DBD), and the E domain (ligand binding-domain; LBD), responsible for ligand binding, nuclear localization, and transcriptional activation (AF-2), are highly conserved with homologies of 90% and 100% respectively, between chicken, man, and rat [11]. In contrast, considerable divergence has occurred in the N-terminal A/B domain, which contains a second activation function, AF-1, the D domain (hinge region), necessary for the maintenance of ER three-dimensional structure [12, 13], and the C-terminal F domain.
ERs have been identified in divergent animal species ranging from nematodes [14, 15] to human [16, 17]. The diverse actions of estrogens in their various target tissues were long thought to be mediated through a single ER, ER (first isolated from human [16]). The presence of a second ER subtype, ERß, has since been demonstrated in mammals, birds, fish, and amphibians (first isolated from rat [18]). Most recently, a third distinct ER, ERß2 (also referred to as ER) has been discovered in teleost fish [19] that is closely related to ERß, suggesting that it reflects a duplification of this gene in part, or all, of the fish phylum. In human, the ER subtype is now designated ESR1 and the ERß subtype is now designated ESR2 (HUGO Gene Nomenclature Committee; www.gene.ucl.ac.uk/nomenclature). In zebrafish (Danio rerio), the ER subtype is now designated Esr1, the ERß1 subtype is now designated Esr2b, and the ERß2 subtype is now designated Esr2a (Official Zebrafish Nomenclature Guidelines; http://zfin.org). Hence, to standardize ER subtype nomenclature, the official protein and gene designations for the zebrafish have been adopted here.
The discovery of multiple ERs has added an additional layer of complexity to our understanding of the action of estrogens and prompted intense interest in the respective role of each subtype. The biological significance of multiple ERs, however, has yet to be fully elucidated. Esr1 and Esr2 are structurally similar, in the rat displaying more than 90% sequence identity in the C domain but only 55% sequence identity in the E domain [18, 20]. They display distinct ligand-binding capacities [20–23] and transcriptional properties [4, 5, 24–28]. Furthermore, they have differential tissue [17, 20, 29–33] and ontogenic [34–36] patterns of expression and they produce distinct knockout phenotypes [37–40]. Moreover, in the mouse, the two different ERs have been reported to have opposite physiological roles in the uterus and prostate [41, 42].
Distinct patterns of expression for esr1 and esr2 have been documented in fish, but these patterns appear to vary between species. In fish, esr1 and esr2 exhibit broad tissue expression but, for the most part, are concentrated in liver and gonad, with considerably lower levels of expression in the brain, pituitary, intestine, and muscle [19, 43–45]. In contrast with these findings, however, esr1 in African catfish (Clarias gariepinus) and esr2a in goldfish (Carassius auratus) have been reported to be expressed predominantly in pituitary and brain, with considerably lower levels of expression in the gonad and liver [46–48]. Although differences in the ligand-binding traits of the ER subtypes have been indicated in fish [49–51], functional differences, in terms of their roles in reproduction and other physiological processes, have not been established in fish.
Over the past two decades, there has been increasing concern about the impacts of chemicals discharged into the environment that possess hormonal activity and disrupt the endocrine systems in wildlife and humans [reviewed in 52– 55]. By far the greatest focus in this work has been on chemicals that mimic estrogens and/or disrupt estrogen signaling pathways. In wild populations of fish living in U.K. rivers, exposure to environmental estrogens induces altered sexual development [55] and this results in fish with reduced fertility [56, 57]. Although the molecular mechanisms of environmental estrogen action are not well understood, it is clear that many of these chemicals function by binding ERs and blocking endogenous estrogen access [58], although additional modes of action [59–61] cannot be discounted.
To fully elucidate the mechanistic pathways by which environmental estrogens are able to modulate endogenous estrogen signaling in fish requires a greater understanding of the distinct roles of the ER subtypes. In this study, the cDNAs for esr1, esr2b, and esr2a were cloned, sequenced, and characterized and their patterns of ontogenic and tissue expression analyzed in relation to development and reproductive status in the fathead minnow (Pimephales promelas), a species used widely by the U.S. Environmental Protection Agency (U.S. EPA) and Organization for Economic Co-operation and Development for research into endocrine disruption. The data generated provide insights into the differential physiological roles of the ER subtypes in fish and provide a foundation for studies on the molecular basis of estrogenic disruption in the fathead minnow.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Fathead minnow embryos were obtained from breeding stocks at the University of Exeter, U.K., and maintained in 5-L glass tanks with constantly flowing (20–40 L/day) dechlorinated water filtered to 5 µm and at a temperature of 25 ± 1°C and under a constant photoperiod regime of 16L:8D. Hatching took place between 4–6 days postfertilization (dpf) (75% hatch at 5 dpf). At 120 dpf, fish were sexed (determined by the presence of secondary sexual characteristics: nuptial tubercles, fatpad and fin coloration/body banding in males; prominence of ovipositor in females) and split into breeding pairs, which were monitored daily for ovulation. Fish were fed to satiation three times daily, with Liquifry No. 1 (Interpet, Dorking, U.K.) and live newly hatched Artemia (cysts supplied by ZM Ltd., Hampshire, U.K.) for fry progressing to TetraMin flake food (Tetra, Melle, Germany) and frozen gamma-irradiated bloodworm (Chironomus sp.; Tropical Marine Center, Chorleywood, U.K.) for adults.
Sample Collection
Fish were killed at incremental time periods during development (5, 10, 20, 40, 80, 120, and 150 dpf) by terminal anesthesia with benzocaine (0.5 g/L; ethyl-p-aminobenzoate; Sigma, Poole, U.K.). Animal use protocols were carried out ethically in accordance with U.K. Home Office guidelines.
Morphometrics and Biochemical Analyses All fish used in the study were measured for fork length (mm) and wet weight (mg; from 40 dpf only). To determine the ontogeny of sexual development in relation to somatic growth in the study population, eight fish were sampled at each time point from 5 to 120 dpf, fixed in toto in Bouin fixative (Raymond A. Lamb Ltd., Eastbourne, U.K.) and transferred into 70% ethanol (Fisher Scientific, Loughborough, U.K.) for histological processing and determination of gonadal maturity (as described in [62]).
To determine reproductive status in fish sampled at 150 dpf, gonads were dissected out and weighed for the determination of gonadosomatic index (GSI = [gonad weight/body weight] x 100) and one gonad from each fish was processed for gonadal histology as described above. A blood sample was collected from the caudal sinus of each fish into chilled heparinized syringes, centrifuged at 13 000 rpm for 5 min and the plasma removed and stored at –80°C until processing. Plasma vitellogenin (VTG) concentration was measured using a carp VTG ELISA validated for use with the fathead minnow [63]. In male fish, the development of secondary sexual characteristics was assessed by removal and weighing of the fatpad for determination of fatpad index (FPI = [fatpad weight/body weight] x 100), determination of tubercle number, and grading of tubercle development (as described in [64]).
Molecular Analyses For determining the ontogeny of expression of fhm esr1, esr2b, and esr2a (from 5 to 120 dpf), eight fish were sampled at each time point. Once fish could be sexed by the presence of secondary sexual characteristics (80 dpf onward), of these eight fish, four were male and four were female. For determining tissue expression of fhm esr1, esr2a, and esr2b (at 150 dpf), eight male and eight female fish were sampled. At 5, 10, and 20 dpf, whole-body samples were collected from individual fish. At 40, 80, and 120 dpf, brain, liver, and gonad tissues were collected from each fish. At 150 dpf, brain, pituitary, gill, liver, gonad, intestine, and muscle tissues were collected from each fish. Upon removal, tissue samples were immediately snap frozen in liquid nitrogen and stored at –80°C until use.
Estradiol Exposure
To determine how the hepatic expression of esr1, esr2b, and esr2a responded to the natural ER ligand E2, adult (120 dpf) male and female fathead minnow (eight males and eight females per treatment) were exposed via the tank water to 100 ng E2/L (Sigma) under flow-through conditions for a period of 14 days. Fish were killed by terminal anesthesia with MS-222 (Sigma) and liver tissues were removed, snap frozen in liquid nitrogen, and stored at –80°C until use.
RNA Extraction
Total RNA was extracted from tissue samples using Tri Reagent (Sigma), following manufacturer's instructions. Total RNA concentration was estimated from absorbance at 260 nm (A260 nm; GeneQuant; Amersham Biosciences, Little Chalfont, U.K.) and RNA quality was verified by electrophoresis on ethidium bromide-stained 1.5% agarose gels and by A260 nm/A280 nm ratios >1.8.
Full-Length Cloning of fhm esr1 and esr2b cDNAs
Full-length fhm esr1 and esr2b cDNA sequences were cloned in three consecutive steps using reverse transcription-polymerase chain reaction (RT-PCR) and rapid amplification of cDNA ends (RACE) strategies. To obtain core partial-length fragments of fhm esr1 and esr2b cDNAs, RT-PCRs were carried out using pairs of oligonucleotide primers (esr1: 5'-AAACTCATCTTTGCTCAGGA-3'/5'-CCTTTGTTGCTCATGTGTCT-3'; esr2b: 5'-TGCTGCTGGCTGGATATTCT-3'/5'-GAGAGCAGCATGAGCAGATG-3'; sense and antisense, respectively; MWG-Biotech, Ebersburg, Germany) designed from conserved regions of esr1 and esr2b cDNAs available in the NCBI GenBank database. Fathead minnow liver cDNA was synthesized from 2 µg RQ1 DNase-treated (Promega, Southampton, U.K.) total RNA using oligo(dT)15 (MWG-Biotech) and Moloney-murine leukemia virus (M-MLV) reverse transcriptase (Promega), following manufacturer's instructions, and PCR was carried out using the puRe Taq Ready-To-Go PCR Bead system (Amersham Biosciences). PCR cycling conditions were as follows: initial denaturation at 94°C for 5 min, followed by 40 cycles of 1 min denaturation at 94°C, 30 sec annealing at 50°C (esr1) or 55°C (esr2b), and 1 min extension at 72°C with 15 min final extension at 72°C. PCR reactions were fractionated by ethidium bromide gel electrophoresis and products of the expected size were ligated into the pCR2.1-TOPO plasmid using the TOPO TA Cloning Kit (Invitrogen) following the manufacturer's instructions. The nucleotide sequences were determined for both sense and antisense strands of the plasmid inserts of two independent clones by automated fluorescence sequencing (Lark Technologies Inc., Saffron Walden, U.K.).
To obtain the 5' and 3' ends of fhm esr1 and esr2b cDNAs, a transcript-specific sense (esr1: 5'-GATGGCTGAGATTTTCGACATG-3'; esr2b: 5'-GGATTGATGTGGAGATCTGTGG-3'; MWG-Biotech; for use in 3' RACE) and antisense primer (esr1: 5'-GTCCAGCATGCACTGCACCATGAAG-3'; esr2b: 5'-AGAGTCCAGCAGCCTCAGAACCTT-3'; MWG-Biotech; for use in 5' RACE) were designed for each transcript to replicate sequence regions of its core fragment sequenced above. Each transcript-specific sense primer was designed upstream of the transcript-specific antisense primer for that transcript, such that 5' and 3' RACE fragments would overlap, enabling construction of full-length cDNAs. 5' RACE was performed with the Smart RACE cDNA Amplification Kit (Clontech Laboratories, Inc., Palo Alto, CA) using fathead minnow liver RNA. 3' RACE was performed using 3'-RACE-ready cDNA for fathead minnow liver, which had been reverse transcribed, as described above, but using a 3' RACE adaptor primer (5'-GCGAGCACAGAATTAATACGACTCACTATAGGTTTTTTTTTTTTVN-3,' where V = A/G/C and N = A/G/C/T; MWG-Biotech), the appropriate transcript-specific sense primer, and a universal antisense, 3' RACE outer, primer (5'-GCGAGCACAGGATTAATACGACT-3'; MWG-Biotech), designed to anneal to the 3' RACE adaptor primer incorporated during reverse transcription. To increase the specificity of the reaction, a nested second round 3' RACE was conducted. 5' and nested 3' RACE products of the expected size were cloned and sequenced as described above. The full-length cDNAs for fhm esr1 and esr2b were assembled by alignment of the appropriate internal core fragment and overlapping 5' and 3' RACE fragments using CLUSTAL W [65].
Partial-Length Cloning of fhm esr2a cDNA
To obtain a core partial-length fragment of fhm esr2a cDNA, a RT-PCR was carried out using a pair of oligonucleotide primers (5'-AACTCCAACATGTGTCTGAGT-3'/5'-CATCTCCAGGAGCAGGTCATA-3'; sense and antisense, respectively; MWG-Biotech) designed from conserved residues of esr2a cDNAs available in the NCBI GenBank database and specific to esr2a subtypes. Reverse transcription and PCR were performed as described above using fathead minnow ovary RNA (with the exception that the annealing temperature during PCR was 52°C) and the product was cloned and sequenced as described above.
Northern Blotting
Twenty micrograms total RNA from female liver tissue was denatured with formaldehyde/formamide using RNA sample loading buffer (Sigma) following the manufacturer's instructions and fractionated by electrophoresis on a 1% agarose gel containing formaldehyde run in 1x MOPS-EDTA-sodium acetate buffer (10x is 0.2 M 3-[N-morpholino]-propanesulphonic acid, 0.01 M EDTA, 0.05 M sodium acetate; pH 7). RNA was blotted onto Hybond XL membranes (Amersham Biosciences) by downward capillary transfer in 5x saline-sodium citrate (SSC) (20x is 3 M sodium chloride, 0.3 M sodium citrate), 10 mM sodium hydroxide, and crosslinked by baking at 70°C for 3 h. Membranes were prehybridized by incubation at 42°C for 30 min in ULTRAhyb buffer (Ambion, Huntingdon, U.K.) and then hybridized overnight at 42°C in the prehybridization buffer containing the 32P-labeled DNA probe. The probes, synthesized by PCR, were 155-base pair (bp), 130-bp, and 264-bp fragments of fhm esr1, esr2b, and esr2a, respectively, which were radiolabeled with 32P using the Rediprime II DNA labeling system and Redivue 32P dCTP (Amersham Biosciences). The membrane was washed twice in 2x SSC, 0.5% SDS at room temperature for 5 min, and then washed twice (esr2a) or three times (esr1, esr2b) in 0.5x SSC, 0.5% SDS at 42°C for 15 min. Hybridized membranes were exposed to Kodak BioMax MS films (Kodak; Rochester, NY) with intensifying screens at –70°C for up to 5 days.
Sequence Analysis
Sequence similarity searches were performed using BLASTn [66] and sequence alignments were carried out using CLUSTAL W [65]. Phylogenetic analysis was done within CLUSTAL X [67] using the Neighbor-Joining method [68] with 1000 bootstrap replicates and trees were drawn using TREEVIEW 1.6.6. [69]. Human progesterone receptor was used as a nuclear receptor outgroup. Sequences from the NCBI GenBank database (www.ncbi.nlm.nih.gov) were used in sequence analysis.
Real-Time PCR
Primers specific for fhm esr1, esr2b, and esr2a were designed with Beacon Designer 3.0 software (Premier Biosoft International, Palo Alto, CA) and purchased from MWG-Biotech (esr1: 5'-CACCCACCAGCCCTCAG-3'/5'-CACCTCACACAGACCAACAC-3'; esr2b: 5'-GCCACCTCCAGATTCAG-3'/5'-CACGACTCTCCACACCTTCAG-3'; esr2a: 5'-GTGACGGATGCTTTGGTATG-3'/5'-GGTGCCTTATGTGGGAGAG-3', sense and antisense, respectively). Primer-pair annealing temperatures were optimized for real-time PCR on a temperature-gradient program. Primer specificity was confirmed by gel electrophoresis, melt curve analysis, and automated fluorescence sequencing of PCR products. To determine the detection range, linearity and real-time PCR amplification efficiency (E; E = 10[–1/slope]) [70] of each primer pair, real-time PCR amplifications were run in triplicate on a 10-fold serial dilution series of fathead minnow liver cDNA, and standard curves were calculated referring the threshold cycle (Ct; the PCR cycle at which fluorescence increased above background levels) to the logarithm of the cDNA dilution.
For the expression studies on fhm esrs, cDNA was synthesized from 1 µg RQ1 DNase-treated total RNA, as described above but using random hexamers (MWG-Biotech). Real-time PCR was performed with the iCycler iQ Real-time Detection System (Bio-Rad Laboratories, Inc., Hercules, CA). Each sample was amplified in triplicate using 96-well optical plates (ABgene, Epsom, U.K.) in a 20-µl reaction volume using 1 µl cDNA, 10 µl 2x Absolute QPCR SYBR Green Flourescein Mix (ABgene), 5 µM of the appropriate forward and reverse primers. Hot start Taq polymerase was activated by an initial denaturation step at 95°C for 15 min followed by 40 cycles of denaturation at 95°C for 10 sec and annealing at 60.5°C (esr1), 61.8°C (esr2a), or 62.5°C (esr2b) for 20 sec and, finally, melt curve analysis. Template-minus and reverse transcriptase-minus negative controls were run for each plate and each sample, respectively. Aliquots of fhm liver cDNA were repeatedly quantified on each plate to assess intra- and interassay variability.
To confirm the integrity of cDNA and to quantify differences in RNA load between samples, esr expression values were normalized to a housekeeping gene. In a preliminary study, the levels of a panel of housekeeping genes (18S ribosomal RNA, ribosomal protein L8, elongation factor 1, gluceraldehyde-6-phosphate dehydrogenase) were measured in each sample. Fhm 18S ribosomal RNA (fhm 18S; NCBI GenBank accession number, AY855349; primers: 5'-AATGTCTGCCCTATCAACTTTC-3'/5'-TGGATGTGGTAGCCGTTTC-3'; annealing at 59°C) and fhm ribosomal protein L8 (fhm rpl8; NCBI GenBank accession number, AY919670; primers: 5'-CTCCGTCTTCAAAGCCCATGT-3'/5'-TCCTTCACGATCCCCTTGATG-3'; annealing at 60°C) exhibited the least variation during development/between tissues and between E2-exposed/nonexposed fish, respectively (Filby and Tyler, unpublished results) and were, therefore, used for these normalizations. Relative expression levels (esr: housekeeping gene) were determined using a development of the arithmetic comparative method (2–
Ct; [71]), which includes a correction for differences in E between the target and housekeeping gene [72]. Results were expressed, however, as relative expression ratios rather than as fold changes from a calibrator sample. The formula, therefore, was as follows:
 |
Data Analysis
Nonlinear regression analysis was performed on the growth data using SigmaPlot 2000 software (Jandel Scientific Software, Rafael, CA). A least-squares method was used to select equation parameters to fit an equation to the data. The equation used for a sigmoidal fit (three-parameter) was y = a/(1 + e–c), where c = x – x0/b. Statistical differences in relative mRNA expression between experimental groups were assessed by one-way ANOVA of log-transformed data, followed by Dunn multiple pairwise comparison test or Student t-test. All statistical analyses were performed using SigmaStat 2.03 software (Jandel Scientific Software). All experimental data are shown as the mean ± SEM. Differences were considered statistically significant at P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Full-length fhm esr1 and esr2b cDNA sequences were obtained in three consecutive steps using RT-PCR and 5'/ 3' RACE strategies. For each transcript, a single core cDNA fragment (esr1: 322 bp; esr2b: 355 bp) was amplified by RT-PCR, which displayed a high nucleotide identity with the respective ER subtype cDNAs of other fish species. Subsequent 5' and 3' RACE amplifications generated 1510 bp (fhm esr1)/1609 bp (fhm esr2b) and 1761 bp (fhm esr1)/908 bp (fhm esr2b) cDNA fragments, respectively. The three overlapping fragments (core, 5' RACE, 3' RACE) for each transcript were then aligned to produce a 3167 bp (fhm esr1) and 2318 bp (fhm esr2b) full-length cDNA sequence for each transcript (submitted to NCBI GenBank, accession numbers AY775183 and AY566178, respectively). Additionally, a partial-length (267 bp) cDNA fragment for fhm esr2a was isolated by RT-PCR, which displayed a high nucleotide identity with esr2a cDNAs of other fish species (submitted to NCBI GenBank, accession number AY929613).
To compensate for the possible errors associated with PCR, two independent PCRs were performed for each ER subtype, including first-strand cDNA preparation and subsequent cloning and sequence analyses. Each full-length cDNA, therefore, was constructed from double-stranded sequencing of six entirely independent clones (two clones of the core fragment; two clones of the 5' RACE fragment; two clones of the 3' RACE fragment). In all cases, independent PCR reactions produced identical sequences. Furthermore, in the regions of overlap between core, 5' RACE, and 3' RACE fragments, we observed a perfect sequence identity for each cDNA. Sizes of the transcripts for fhm esr1 and esr2b determined by sequencing were confirmed by Northern blot analysis, with single transcripts in the region of 3–3.3 kb and 2.3–2.4 kb, respectively, present in liver from female fhm (Fig. 1). Northern blot analysis additionally revealed that the size of the mRNA transcript for fhm esr2a was approximately 2.5–2.8 kb (Fig. 1).
|
The coding regions of fhm esr1 (nucleotides 189–1871) and esr2b (nucleotides 215–1978) cDNAs encoded putative Esr peptides of 560 and 587 amino acids, respectively. The partial-length fhm esr2a fragment encoded an 89 amino acid portion of the putative fhm Esr protein corresponding to the 3' end of the E domain. Amino acid identity was only 43% between fhm Esr1 and fhm Esr2b, 52% between fhm Esr1 and the partial fhm Esr2a, and 70% between fhm Esr2b and the partial fhm Esr2a, and amino acid differences were evident throughout their coding regions (Fig. 2), indicating that the three fhm ER proteins were generated from distinct genes. Fhm Esr1 had a 47% identity with human ESR1 and a lower (44%) identity with human ESR2. Fhm Esr2b had only a 40% identity with human ESR1 but a 53% identity with human ESR2. The partial sequence for fhm Esr2a had a 47% identity with the corresponding region of human ESR1 but a 55% identity with the corresponding region of human ESR2. These data indicate that fhm Esr1 derives from an Esr1 ancestral subtype and that fhm Esr2a and Esr2b derive from an ancestral Esr2 subtype, and this was confirmed by phylogenetic analysis (Fig. 3). The phylogenetic tree revealed a clear division between teleost ER and tetrapod ER sister clades within each major ER subtype. Inside the teleost Esr1 clade, fhm Esr1 had the highest similarity to zebrafish Esr1 (88% similarity). Inside the teleost Esr2 clade, fhm Esr2b had the highest similarities with zebrafish Esr2b, goldfish ERß2, and Taiwan shoveljaw carp ER (85% similarity), and the partial fhm Esr2a had the highest similarity with the same region of rainbow trout ERß (100% similarity). According to the system of synapomorphies observed to be diagnostic of the teleost Esr2a clade versus the Esr2b clade [19], fhm Esr2a would be classified as an Esr2a rather than an Esr2b and fhm Esr2b would be classified as an Esr2b rather than as an Esr2a.
|
|
Based on an amino acid alignment of fhm Esr1 and Esr2b with human ESR1 and ESR2, functional domains (as defined in [10]) were designated to fhm ER residues as follows: A/B domain: residues 1–154 (Esr1) and 1–171 (Esr2b); C domain: residues 155–220 (Esr1) and 172–237 (Esr2b); D domain: residues 221–319 (Esr1) and 238–342 (Esr2b); E domain: residues 320–518 (Esr1) and 343–541 (Esr2b); F domain: residues 519–560 (Esr1) and 542–587 (Esr2b) (Fig. 2). There was a high amino acid identity of the C and E domains with human ERs (93–96% and 61– 70%, respectively; Fig. 2). In addition, a high degree of conservation of residues within established functional motifs was evident. For example, within the C domain, the positions of eight cysteine residues, which constitute the two zinc-finger motifs (CI and CII; [73, 74]) and the P and D boxes [75, 76] were conserved in fhm ERs. Furthermore, within the E domain, four residues identified as being critical for ligand binding (Gly521, His524, Leu525, Met528; [77] located close to Cys530) and ligand-dependent transcriptional activity (Met517Leu and Tyr526Ser; residue positions referring to human ESR1 [77, 78]) (Fig. 2) were conserved in all three fhm ERs. However, various other substitutions between fhm Esr1 and fhm Esr2b/2a were noted within this region (Gln504Leu; Gln506His; Leu509Met) and a single amino acid substitution (Gln547Asn; residue position according to human ESR1) between fhm Esr1 and Esr2b was observed within the ligand-dependent transcriptional activation function, AF-2. Low sequence identities between fhm ERs and human ERs were observed in the A/B (11–24%), D (30–47%,) and F (10–20%) domains (Fig. 2).
Development of Real-Time PCR Assays for fhm esr1, esr2b, and esr2a
Real-time PCR assays for quantification of fhm esr1, esr2a, and esr2b and for the housekeeping genes fhm 18S and fhm rpl8 had detection ranges of at least five orders of magnitude (data not shown). Specificity of primer sets throughout this range of detection was confirmed by the observation of single amplification products of the expected size and Tm and sequence (data not shown). All assays were quantitative, with standard curve (mean Ct versus log cDNA dilution) slopes of between –3.03 and –3.34, translating to high E values of 2.00–2.14. Over the detection range, the linear correlation (R2) between the mean Ct and the logarithm of the cDNA dilution was >0.98 in each case. Assays had a high level of precision and reproducibility with intraassay coefficient of variation (CV) of 2.42%; (n = 96) and interassay CV (n = 9) of 3.73% (fhm esr1), 5.26% (fhm esr2b), 2.86% (fhm esr2a), 3.01% (fhm rpl8), and 4.16% (fhm 18S).
Expression of fhm esr1, esr2b, and esr2a
Ontogeny of sexual development in the study population We employed gonadal histology and measurements of somatic growth to enable us to link the stages of sexual development with the expression patterns of fhm ERs. Regression analysis on fork length against age (Fig. 4A) using the sigmoidal model produced R2 values of 0.99 for all groups. At 40 dpf, fish wet weight was 0.405 ± 0.021 g; at 80 dpf (when sex was discernable from secondary sex characteristics), 2.228 ± 0.208 g (males) and 0.910 ± 0.026 g (females); at 120 dpf, 2.790 ± 0.205 g (males) and 1.020 ± 0.056 g (females); and at 150 dpf, 3.206 ± 0.105 g (males) and 1.308 ± 0.061 g (females). At 150 dpf, GSI was 1.330 ± 0.070 (males) and 12.380 ± 0.795 (females) and plasma VTG concentrations were 1.590 ± 0.753 µg/ ml (males) and 182.083 ± 19.567 µg/ml (females). At 150 dpf, male fish had an FPI of 3.82 ± 0.306, 16.5 ± 0.793 tubercles and a tubercle prominence index [61] of 3.833 ± 0.112. At 150 dpf, all female fish were regularly ovulating with an interval of 4.334 ± 0.181 days.
|
The timing of sex-cell differentiation and progression of gonadal development (assessed by histology), in relation to age, is presented in Figure 4B. On the day of hatch (5 dpf), the gonadal analges had not yet formed but, by 10 dpf, the gonads consisted of a few primordial germ cells covered by a thin layer of somatic cells. Differentiation into male or female gonad, however, did not occur until 20 dpf. At this time, the male gonad was characterized by somatic cells dispersed among the germ cells and the gonad had a single point of attachment to the peritoneal wall. In contrast, the female gonad was characterized by two points of attachment to the peritoneal wall, and meiotic germ cells were located in the center of the gonad with somatic cells restricted to the periphery. The ovary was also surrounded by an ovarian cavity. At 40 dpf, ovaries contained germ cells of all stages up to primary oocytes. At 80 dpf, ovaries also contained secondary oocytes (including vitellogenic oocytes), which increased in prevalence with growth of the gonad to 120 dpf (data not shown). There were no further changes in the size or germ cell composition of ovaries between 120 and 150 dpf. Testes remained quiescent until 80 dpf, when testis lobules had developed and all stages of germ cells were apparent but were predominantly spermatogonia and spermatocytes. At 120 dpf, testes were fully mature, with lobules filled with spermatozoa. The only apparent difference between males at 120 and 150 dpf was an overall increase in size of the testes (data not shown).
Expression of fhm esr1, esr2b, and esr2a During Development
Expression of fhm esr1, esr2b, and esr2a was detected at all stages of development studied. Whole-body expression of fhm esr1 (one-way ANOVA; F2,21 = 7.993; P < 0.005), esr2b (one-way ANOVA; F2,21 = 21.078; P < 0.001), and esr2a (one-way ANOVA; F2,21 = 9.165; P < 0.001) varied between 5 and 20 dpf (data not shown). In whole-body extracts during all early life stages (5–20 dpf), fhm ER mRNAs were expressed to different levels, in the order esr1 < esr2b < esr2a (data not shown).
Between 40 and 120 dpf, fhm esr1, esr2b, and esr2a were measured in brain, gonad, and liver tissues of individual fish, with the sex of these fish identifiable from 80 dpf onward (Fig. 5). In liver, expression of fhm esr1 increased between 40 and 80 dpf but increased further between 80 and 120 dpf only in females. Liver fhm esr2b expression similarly increased between 40 and 80 dpf but with a lower proportional increase over these developmental stages compared with fhm esr1 (fhm esr1: 20-fold increase versus fhm esr2b: 4-fold increase). Expression of fhm esr2a in the liver did not vary between these developmental stages. In testis, there was no change in expression of fhm esr1 between 40 dpf and 80 dpf but a 4-fold increase in ovary. No change in expression of fhm esr1 occurred in gonad between 80 and 120 dpf for either sex. At both 80 and 120 dpf, there was sexual dimorphic gonadal expression of fhm esr1; expression in ovary was approximately 6.5-fold higher than in testis. A progressive increase in gonadal expression occurred for fhm esr2b between 40 and 120 dpf, both in males and females, although the degree of increase was smaller for females than for males. At 120 dpf, fhm esr2b was more highly expressed in testis than ovary. Expression of fhm esr2a in the gonad was significantly higher at 40 dpf than at any other developmental stages. There were no statistically significant changes in the expression of fhm esr1, esr2b, or esr2a in the brain between 40 and 120 dpf.
|
Expression of fhm esr1, esr2b, and esr2a in Adult Body Tissues
In adult fish (150 dpf), expression of fhm esr1, esr2b, and esr2a occurred in all body tissues with the exception of fhm esr1, which was not detected in muscle (Fig. 6). The levels of expression for fhm esr1 (one-way ANOVA; F11,82 = 24.410; P < 0.001), esr2b (one-way ANOVA; F13,96 = 11.183; P < 0.001) and esr2a (one-way ANOVA; F13,96 = 6.045; P < 0.001), however, differed between tissues.
|
The lowest level of expression of fhm esr1 occurred in intestine and the highest in liver (expression in liver was approximately 40-fold higher than in intestine). There was similarly a high level of expression of fhm esr1 in pituitary (approximately 15-fold higher than in intestine) and in brain and gonad (approximately 5-fold higher than in intestine). Low-level expression of fhm esr1 occurred in gill. Fhm esr2b had the lowest level of expression in brain and, in parallel with fhm esr1, was most highly expressed in liver, where expression was approximately 18-fold higher than in brain. Fhm esr2b also had a high level of expression in intestine (approximately 9-fold higher than brain), pituitary (approximately 5-fold higher than brain), and gonad (approximately 4-fold higher than brain). Lower levels of expression of fhm esr2b were found in gill and muscle. In contrast with fhm esr1 and esr2b, fhm esr2a had the lowest level of expression in the liver, where expression was approximately 30-fold lower than in the intestine (the tissue with the highest expression). High-level expression of fhm esr2a also occurred in the pituitary, muscle, and brain. There was a sexually dimorphic expression of fhm esr1 in gonad (Student t-test; t = –3.223; df = 14; P < 0.01) and pituitary (Student t-test; t = 2.464; df = 14; P < 0.05) and for fhm esr2b in pituitary (Student t-test; t = 0.640; df = 14; P = 0.01) and liver (Student t-test; t = 2.780; df = 14; P = 0.01).
Fhm esr1 had a significantly higher level of expression than esr2b in liver (Student t-test; t = 3.869; df = 30; P < 0.001) and brain (Student t-test; t = 4.273; df = 26; P < 0.001). In contrast, fhm esr2b was more highly expressed than fhm esr1 in intestine (Student t-test; t = –8.030; df = 30; P < 0.001). There were no significant differences between the levels of expression of fhm esr1 and fhm esr2b in gonad, pituitary, or gill. Fhm esr2a had a significantly lower level of expression than fhm esr1 (Student t-test; t = –12.681; df = 30; P < 0.001) and esr2b (Student t-test; t = –6.791; df = 30; P < 0.001) in the liver but was expressed at a higher level than fhm esr1 and esr2b in all other tissues examined, with the exception of gill tissue.
Expression of fhm esr1, esr2b, and esr2a Following Exposure to E2
To determine how the fhm ERs were regulated by the natural ligand E2, we exposed adult male and female fish to 100 ng E2/L for 14 days and measured esr1, esr2b, and esr2a expression in the liver and compared this with expression in untreated fish. There was a significant induction (approximately 5-fold) of fhm esr1 in the liver of male fish exposed to E2 (Student t-test; t = –2.235; df = 15; P = 0.041). There appeared to be a similar induction of esr1 in the liver of female fish, but this difference was not statistically significant (Fig. 7). Levels of fhm esr2b and esr2a remained unchanged following 14 days of exposure to E2 (Fig. 7).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
sequence of Wilson and colleagues ([79]; NCBI GenBank accession number: AY727528). Our isolation of two distinct esrs from the fathead minnow is in keeping with findings in several other teleosts [19, 43, 45, 46, 80, 81]. Examination of the predicted amino acid sequences for fhm Esr1 and Esr2b revealed the highest levels of identity between fhm ER subtypes in the C and E domains, consistent with their crucial roles in DNA binding and ligand binding, respectively. In the C domain, 61 of 66 amino acids, including the P and D boxes, which interact directly with regulatory regions of DNA [75, 76], were identical in fhm Esr1 and Esr2b, as in human ESR1 and ESR2. This indicates that the two ERs may bind to the same type of response elements and with the same, or similar, affinities. Differences in the abilities of ER subtypes to activate particular response elements [25] and different affinities for the same response elements [21, 82], however, have been described in mammals.
The high conservation between the E domains of fhm ERs also implies similar binding properties of ER ligands for the ER subtypes. However, in mammals, single amino-acid substitutions in the E domain are responsible for the differences in binding affinity to ERs [83], and even single residue mutations in this domain can dramatically alter ligand binding and transactivation characteristics [77, 84, 85]. In human ERs, estrogen binding involves residues close to Cys530 and the four residues identified to be particularly important (Gly521, His524, Leu525, Met528 [77]) are conserved between fhm ERs. However, substitutions were observed further toward the N-terminal or C-terminal of this region, which, in human and mouse, have effects on receptor dimerization (Gln504Leu, Gln506His, Leu509Met [86, 87]) and hormone-dependent transactivation (Met517Leu, Tyr526Ser; residue positions referring to human ESR1 [77]). Furthermore, substitutions were seen between fhm ERs in all four residues at positions surrounding the human ESR1 binding pocket (Leu349, Met421, Tyr526, Cys530), in accordance with the scheme presented by Hawkins and Thomas for teleosts [88], which are believed to account for differences in ligand-binding properties observed for Atlantic croaker ERs [88]. Lastly, a single amino acid substitution between fhm Esr1 and Esr2b (Gln547Asn; residue position according to human ESR1) was observed within the ligand-dependent transcriptional activation function, AF-2. It is very probable, therefore, that fhm ERs may have differential ligand-binding preferences and ligand-dependent transactivation properties, as has been demonstrated for ERs in other fish species [49–51]. Future functional studies of fhm ERs are, however, required to characterize any differences in functional characteristics between ER subtypes in this species.
The observation of very low identities between fhm Esr1 and Esr2b in the A/B, D, and F domains is in keeping with the hypervariability known to occur generally in these domains. Notably, fhm Esr1 had a considerably shorter A/B domain than fhm Esr2b (154 compared with 171 amino acids), which is consistent with other teleosts [46, 47, 89]. However, the existence of multiple teleost Esr1 isoforms, with both longer and shorter A/B domains, appears to be common [49, 90, 91]. Differential A/B domain length between ER subtypes can lead to differences in ligand-independent transactivation (AF-1) function [90]. Indeed, in mammals, the A/B domain has been identified as the site for the main functional differences between ER subtypes [45].
Currently, there is a paucity of information regarding the expression and respective roles of ER subtypes in fish, both during development and in adults. Furthermore, ER expression studies in fish have principally applied qualitative and semiquantitative Northern blot and RT-PCR techniques [e.g., 45, 46, 89, 92]. In this study, real-time PCR assays were developed and validated to quantify fhm esr1, esr2b, and esr2a mRNAs during development to sexual maturity and in various body tissues in the fathead minnow.
All three fhm ERs were expressed during early (5–20 dpf) development. However, future studies to localize this expression within tissue types are required to better identify the possible roles of ERs during these critical stages of sexual differentiation in fathead minnow. In maturing (40– 120 dpf) and adult (150 dpf) fhm, esr1, esr2b, and esr2a had broad patterns of tissue expression, consistent with the pleiotropic roles of estrogens. Expression of esr1 was not, however, detected in muscle in fathead minnow, which differs from findings in other fish [47, 48, 89] and in mammals [93–95], where both ERs are expressed.
The highest level of tissue expression of both fhm esr1 and esr2b (up to 40-fold and 18-fold higher, respectively), occurred in the liver, the site of synthesis of the yolk protein precursor, VTG, which is stimulated by E2 and mediated through the hepatic ER [96]. Although the high levels of both fhm esr1 and esr2b in the liver suggest that E2-dependent transcriptional regulation of the VTG gene(s) may be associated with the coordinated role of both ER subtypes, recent data suggest that esr1 plays the dominant role and that esr2b may have no function in this process [80, 97, 98]. In support of this, hepatic expression of fhm esr1 increased progressively with sexual development in females, reflecting the role of E2 in promoting vitellogenesis. It is well established that male fish are also responsive to VTG induction on exposure to estrogens [99] but, due to the negligible circulating E2 levels in male fish, hepatic esr1 mRNA expression is usually considerably lower or undetectable in males compared with in females [48, 81, 90, 100–102]. In fathead minnow, we found relatively high levels of hepatic esr1 in males (as has been reported in some other fish species [47, 49, 89, 103]), but the physiological significance of this is not known. The concentration of VTG (as a biomarker of exposure to exogenous estrogen [99]) in these male fish was negligible and excluded the possibility that this finding resulted from inadvertent exposure to exogenous estrogen. The significance of high fhm esr2b mRNA levels in the liver is also difficult to interpret, given that teleost ER-ß1 appears not to have a role in the regulation of vitellogenesis. The mRNA levels are not necessarily indicative of protein levels, however, and thus further protein analysis is required to confirm that these mRNAs are translated into ER protein in these tissues. Fhm esr2a was expressed at far lower levels than fhm esr1 and esr2b in the liver and did not vary in the liver during sexual development, which is in keeping with data from other fish species [80, 97] and suggests that its presence in the liver reflects hepatic roles divergent from that of fhm esr1 and fhm esr2b.
The role of ER subtypes in the gonads of fish is not well understood. The traditional view of E2 as the female hormone has been challenged in recent years [98, 104] and the presence of estrogen, ER, and aromatase in the testis is now well recognized [reviewed in 105]. In fish testis, ER is known to be autoregulated by E2 [103], but contradictions in the reported localizations of ER subtypes in this tissue have limited our understanding of their roles in both mammals and lower vertebrates. The higher testis expression of fhm esr2b and esr2a compared with fhm esr1 and the upregulation of fhm esr2b but not fhm esr1 during the key period of testicular differentiation implies that esr2 subtypes may play the main role in ER-mediated testicular function and spermatogenesis in fish. Interestingly, in mammals, Esr2 is the predominant, and potentially the only, ER in germ cells [105]. In females, estrogens have long been known to have significant impacts on ovarian functions. Studies in mammals have suggested that E2 may upregulate follicle-stimulating hormone receptors within the ovary [106] and stimulate the release of growth factors that promote the growth and development of the granulosa and theca cells that surround the oocyte [106, 107]. In fathead minnow, fhm esr2a, rather than fhm esr1 or esr2b, was the dominant ER subtype in the ovary, suggesting that ER subtypes have different and sex-specific roles in reproduction. In several fish species, data are available on the localizations of ER protein and mRNA in the testis and ovary [e.g., 108–109], and future experiments localizing ER mRNA and protein in the gonad should help to better define their respective roles in the fathead minnow in sexual development and function.
The expression of fhm ERs in the brain-pituitary complex, although at low levels, agrees with the known effects of estrogens on the central nervous system, notably in the regulation of luteinizing hormone (LH) synthesis and release from pituitary gonadotrophs [reviewed in 110–112]. The specific functions of the different ER subtypes in mediating these processes have not been distinguished but ER subtypes have both distinct distributions and colocalizations in the brain and pituitary of fish [19, 45] and mammals [113, 114], suggesting both distinct and common roles. Indeed, studies on ER knockout mice have shown that both types of ER are required for E2-mediated regulation of LH [115]. In our studies on fathead minnow, there were no changes in the levels of expression of fhm ERs in the brain during the period of sexual differentiation. Due to the small size of the fathead minnow, we were unable to dissect out the pituitary until fish had reached their maximal size and, therefore, we were unable to measure pituitary ER mRNA levels during sexual development. Evidence from other fish species, however, has shown that pituitary ER expression increases during pubertal development, in concert with LHß subunit mRNA levels [48, 81]. It is interesting to note that expression of ER in the brain appeared to be independent of sex throughout the period of sexual development, as has been observed in other fish species [47, 92]. In contrast, in the pituitary of sexually mature fish, both fhm esr1 and esr2b were expressed more highly in males compared with females, possibly reflecting sex-specific differences in the control of the LH-release mechanisms local to the pituitary. As with gonadal esr expression, further studies localizing the expression of esrs within the brain-pituitary complex of the fathead minnow are likely to aid in identifying the respective roles of ER subtypes in these tissues.
The expression of all fhm ERs in the intestine and gills, tissues that historically have not been designated as estrogen-responsive targets, has also been observed in other fish species (detection of ER in intestine [44–49, 89, 92, 103], detection of ER in gill [46, 49, 89, 92]). Intriguingly, we found very high expression of fhm esr2s, in particular esr2a, but not fhm esr1 in intestine. This again has been observed in other fish species [44, 45, 47, 81, 103], but its possible significance has not previously received comment. In mammals, both ERs are present in intestine [116, 117] and estrogen has been implicated in the regulation of intestinal function [117–120]. In salmonids, it has been suggested [89, 92] that expression of ER in gill and intestine may be connected with a role for estrogen in osmoregulation because E2 and various xenoestrogens antagonize morphological, metabolic, and physiological changes that occur during smoltification [121–124].
Interestingly, exposure of adult fathead minnow to the natural ER ligand E2 invoked differential responses of fhm esr genes in the liver. Our data for esr 1 complies (at least in males) with the established information in other oviparaous species [e.g., 80, 97, 125–128], with an upregulation of the hepatic esr1 gene in response to E2. The absence of a significant upregulation of hepatic fhm esr1 in female fathead minnow was likely to be due to the fact that hepatic esr1 expression was already high in these fish and probably close to its maximum level of expression (associated with the high circulating plasma E2 associated with vitellogenesis). In contrast, the hepatic expression of both fhm esr2b and esr2a genes were not significantly affected by E2. Similar findings have been observed for zebrafish hepatic esr2s, which were either not affected or only slightly affected by E2 [97]. In largemouth bass, hepatic esr2b levels were not significantly affected, but esr2a levels were upregulated by E2 [80]. Further experiments are required to better define these differences in induction of the fhm ER subtypes with multiple doses of E2. However, these data presented support the theories that Esr2s may not be involved in the regulation of vitellogenesis [80].
Our work has demonstrated that, in fathead minnow, Esr1 and Esr2 subtypes have differences in their biochemical structures, indicating different functional characteristics, and differences in their tissue and ontogenic patterns of expression and inducibility by E2, implying distinct physiological functions. While considerable overlap was observed in the tissue distributions of fhm esr1, esr2b, and esr2a, there is evidence from other species that, within a particular tissue, their expression may not be within the same cell population [e.g., 19, 45, 113, 114], suggesting that their functions in these tissues are not redundant. In vivo studies of ER-knockout mice have demonstrated that, while Esr1 is critical for fertility in both sexes, Esr2 is not and, therefore, may not be required for some aspects of reproduction [1, 38, 39]. Clearly, further studies are required to demonstrate whether this is also the case in fish. In light of recent evidence that some environmental estrogens exhibit preferential binding for a particular ER subtype in fish [129], as in mammals [20, 22, 130], it is critical that we develop a more comprehensive understanding of the respective roles of each ER subtype if we are to provide mechanistic explanations for the effects of exposure to these chemicals in fish.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: FAX: 44 1392-263700; a.l.filby{at}exeter.ac.uk
Received: 6 January 2005.
First decision: 3 February 2005.
Accepted: 26 May 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and ERß at AP1 sites. Science 1997 277:1508-1510/ß)-dependent activation at GC-rich (Sp1) promoter elements. J Biol Chem 2000 275:5379-5387 and ß. Endocrinology 1997 138:863-870 and estrogen receptor ß to partial estrogen agonists/antagonists. Mol Pharmacol 1998 54:105-112 or estrogen receptor-ß. Endocrinology 1999 140:800-804 and ß reveals differential specificity in target promoter regulation by xenoestrogens. J Biol Chem 2002 277:44455-44461 and -ß mRNA in the rat central nervous system. J Comp Neurol 1997 388:507-525[CrossRef][Medline] mRNA expression in male and female rhesus macaques. Mol Cell Endocrinol 1998 146:59-68[CrossRef][Medline]
