Communicated by: Hiroshi Handa
Chromatin immunoprecipitation-mediated target identification proved aquaporin 5 is regulated directly by estrogen in the uterus
Article first published online: 14 AUG 2006
© 2006 The Authors. Journal compilation © 2006 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd
Genes to Cells
Volume 11, Issue 10, pages 1133–1143, October 2006
How to Cite
Kobayashi, M., Takahashi, E., Miyagawa, S.-i., Watanabe, H. and Iguchi, T. (2006), Chromatin immunoprecipitation-mediated target identification proved aquaporin 5 is regulated directly by estrogen in the uterus. Genes to Cells, 11: 1133–1143. doi: 10.1111/j.1365-2443.2006.01009.x
- Issue published online: 14 AUG 2006
- Article first published online: 14 AUG 2006
- Received: 10 April 2006Accepted: 17 May 2006
Estrogens play a central role in the reproduction of vertebrates and affect a variety of biological processes. The major target molecules of estrogens are nuclear estrogen receptors (ERs), which have been studied extensively at the molecular level. In contrast, our knowledge of the genes that are regulated directly by ERs remains limited, especially at the level of the whole organism rather than cultured cells. In order to identify genes that are regulated directly by ERs in vivo, we used estrogen treated mouse uterus and performed chromatin immunoprecipitation. Sequence analysis of a precipitated DNA fragment enabled alignment with the mouse genomic sequence and revealed that the promoter region of the gene encoding aquaporin 5 (AQP5) was precipitated with antibody against ERα. Quantitative PCR and DNA microarray analyses confirmed that AQP5 is activated soon after administration of estrogen. In addition, the promoter region of AQP5 contained a functional estrogen response element that was activated directly by estrogen. Although several AQP genes are expressed in the uterus, only direct activation of AQP5 could be detected following treatment with estrogen. This chromatin immunopreciptation-mediated target identification may be applicable to the study of other transcription factor networks.
Estrogens play crucial roles in reproduction and other biological processes via two pathways. Firstly, they exert their effects genomically through the estrogen receptors (ERs), which function as ligand-dependent transcription factors that bind DNA at conserved estrogen response elements (EREs). Two ERs (ERα and ERβ) have been identified in mammals and since ERα knockout mice are sterile, the former is considered to play an essential role in reproduction (Lubahn et al. 1993). Secondly, estrogens function in a non-genomic manner by rapid activation of membrane-initiated kinase cascades such as the MAPK (Migliaccio et al. 1996) and phosphatidylinositol-3-OH kinase (Simoncini et al. 2000) signaling pathways. As the net effect of estrogen represents the integration of both genomic and non-genomic activities (Kahlert et al. 2000; Wong et al. 2002), it is important to distinguish between these pathways when attempting to understand its functions.
Membrane binding assays (Inoue et al. 1993), computational analyses (Bourdeau et al. 2004; Kamalakaran et al. 2005) and chromatin immunoprecipitation (ChIP) (Carroll et al. 2005; Laganiere et al. 2005) have been used to identify ER target genes in a non-biased manner. In particular, the recently developed ChIP technique has enabled investigation of genes at a genome-wide level. However, these approaches have been applied primarily to cultured cell lines such as MCF-7. Consequently, the estrogen target genes of whole organisms remain poorly understood in comparison to cultured cell lines.
The uterus is known to be a major target organ for estrogen and its condition both physiologically and morphologically depends upon estrogen levels. ERα is expressed at high levels in the uterus and in ERα knock-out mice the effects of estrogen are canceled. From studies using ERα and ERβ null mice (Lubahn et al. 1993; Krege et al. 1998), it is known that the former plays an essential role in reproduction.
Although the influence of estrogen on the uterus is both well recognized and studied, our understanding of estrogen-activated transcriptional networks remains limited. Several groups have examined the uterine gene expression profile (Watanabe et al. 2002; Hewitt et al. 2003; Moggs et al. 2004), but none have determined whether these genes were activated directly by ER. Thus, it is not clear to what extent these genes are activated as a direct result of genomic action by estrogen in the target organ.
One of the prominent effects of estrogen is water imbibition in the uterine endometrium. This estrogen-stimulated water transport is considered to be important in the periimplantation period to change the uterine environment. For the regulation of water transportation, water channels termed aquaporins (AQP) play a critical role. AQPs are a family of transmembrane channel proteins that have six transmembrane domains. Thirteen AQPs (AQP0–AQP12) have been identified in mammals and the expression of AQPs in the uterus has also been studied at the protein and RNA level (Offenberg et al. 2000; Jablonski et al. 2003; Richard et al. 2003), although there has not been a study to address whether the AQP genes could be directly activated by estrogen.
In order to identify the genes that are regulated directly by estrogen in the whole organism, we used ChIP to select activated genes in the mouse uterus. A DNA fragment that was recognized by the ER was precipitated and identified as the 5′ region flanking AQP5. We analyzed its function and response to estrogen as well as the responses of other AQP genes.
AQP5 is a target of ERα
We used ChIP to investigate the uterine genes targeted by ERα. The mouse uterus was fixed 1 h after estrogen treatment, after which genomic DNA was fragmented and precipitated using anti-ERα antibody. This DNA was amplified by PCR and sequenced and compared to the mouse genome. One of the sequences of a precipitated DNA fragment mapped to Chr:15 99647562-99647991, located only 6 bases upstream of AQP5 (Chr:15 99,647,997-99 652 041; Fig. 1). In contrast, no specific DNA fragments that could be assigned to mouse genome sequence were recovered from ChIP of the untreated uterus (data not shown).
Although a canonical ERE was not identified in the precipitated sequence, a highly conserved motif was found upstream of AQP5 (between −85 and −73; Fig. 1). To confirm binding of ERα to the putative ERE, we performed a conventional ChIP with fragments from the 5′ flanking and coding regions of AQP5 (Fig. 2A and Table 1). Only fragments containing the putative ERE and its adjacent sequences could be precipitated following administration of estrogen (Fig. 2B). The other fragments could not be precipitated with anti-ERα antibody, even after administration of E2. Similar results could be obtained with monoclonal anti-ERα antibodies. These results suggest that the putative ERE plays a key role in the estrogen response. In contrast, all DNA fragments could be precipitated with anti-acetylated histone H3 antibody (anti-AcH3). It is known that the modification states of the N-terminal tails of hitones H3 and H4 play important role in chromatin formation (Richards & Elgin 2002) and the fact that Lys9 of H3 was acetylated indicated that the chromatin state of the 5′ flanking region of AQP5 was active.
|Primer set||Position||Direction||Primer sequence|
We examined the temporal changes in binding between ERα and the putative ERE following administration of estrogen. Binding could be detected 1 h after estrogen administration and continued for at least 6 h (Fig. 2C). In contrast, no PCR product was detected from samples precipitated with IgG (data not shown).
Identification of an ERE in the 5′ flanking region of AQP5
In order to determine whether the putative ERE plays a functional role in estrogen-dependent transcriptional activation, we used three luciferase reporter constructs (p611AQP5-luc, p103AQP5-luc and p13AQP5-luc) containing deletions of the 5′ flanking region of AQP5 (−611, −103 and −13 to +74, respectively).
Only the positive control (containing 3 × canonical ERE) and the two constructs containing the putative ERE (p611AQP5-luc and p103AQP5-luc) were activated by estrogen, whereas the construct that did not contain the putative ERE (p13AQP5-luc) and a construct in which the element was mutated (pm103AQP5-luc; GGTCAaaTCA), were unaffected (Fig. 3). These results suggest that there is a functional ERE element in the 5′ region of AQP5 that is regulated directly by ERα.
DNA microarray analysis
In order to obtain a uterine gene expression profile following treatment with estrogen, and to identify gene targets that were regulated directly by ERα rather than via secondary effects of estrogen, we performed a DNA microarray analysis of mouse uterus isolated 1 h following treatment. At 1 h following administration of estrogen, we identified a more than twofold up- or down-regulation in 216 and 204 genes, respectively. As AQP5 is located between AQP2 and AQP6, we examined whether other AQP genes were affected by estrogen. At 1 h post-treatment, expression of AQP5 and AQP8 was activated, whereas expression of the AQP genes in the proximity of the former (AQP2 and AQP6) was unaffected (Fig. 4).
Quantitative PCR analysis of AQP genes
As the DNA microarray indicated early activation of AQP5 and AQP8 by estrogen, we examined temporal expression changes of all AQP genes using quantitative PCR, because it has been reported several AQP genes are expressed in uterus and spatiotemporally regulated during the periimplantation (Richard et al. 2003). We confirmed that AQP5 and AQP8 were activated within 1 h of estrogen administration and that activation continued for at least 2 h. Although AQP5 activation declined after 6 h relative to untreated samples, the levels of transcript remained elevated at 24 h (Fig. 5). The other AQP genes, including AQP2 and AQP6, were not activated by estrogen.
Although uterine expression of AQP genes has been examined by other groups at the protein and RNA levels, not all of these observations were in agreement. Weak ERE motifs were found in a proximal promoter region (< 1 kb) upstream of AQP8 and similar motifs were identified upstream of other AQP genes (Table 2). Thus, we examined whether the motifs upstream of the various AQP genes were functional EREs.
|Gene||Position||Putative ERE sequence||Score||Position||Direction||ChIP primer sequence|
Using a reporter system, we examined whether the regions flanking the AQP genes could respond to estrogen stimulation. We prepared transcriptional fusions comprising 1 kb of sequence upstream of each AQP gene and the gene encoding luciferase. With the exception of AQP5, no AQP genes were activated by estrogen (Fig. 6), including AQP8, which had been activated by the administration of estrogen in earlier experiments (Figs 4 and 5).
Next, in order to examine if weak EREs might be recognized by ERα in the uterus, we performed a chromatin immunoprecipitation using anti-ERα. PCR primers were designed to amplify putative ERE sequences in the 5′ flanking region of each AQP gene (Table 2). As the only DNA fragment that could be amplified by PCR contained the ERE of AQP5 (Fig. 7), these results suggest that ERα cannot bind to the flanking region of other AQP genes, although some other AQPs have half ERE (AQP2,3,7, and 12).
In this study, we identified AQP5 as an ERα target gene through a combination of ChIP and DNA microarray analysis. Previously, we identified hundreds of genes in the mouse uterus that were activated by exposure to estrogen (Watanabe et al. 2002), but it was not clear whether those genes were activated directly by estrogen via ERα. We have identified a direct target of ERα using chromatin immunoprecipitation and DNA microarray analysis, and this methodology can be applied to other transcription factors, including other nuclear receptors.
Identification of the target genes of transcription factors is critical to our understanding of transcriptional networks. For example, as canonical EREs were rarely found in promoter regions (O’Lone et al. 2004), it was difficult to identify the target genes of ERs by sequence motifs alone; filter binding (Inoue et al. 1993), computational approaches (Bourdeau et al. 2004; Kamalakaran et al. 2005) and ChIP on Chip (Carroll et al. 2005; Laganiere et al. 2005) have been applied to their identification. As these studies focused primarily on cultured cells, no information regarding whole organisms was available. Previously, we demonstrated the differential expression of estrogen response genes between tissues and throughout various stages of maturation (Watanabe et al. 2004). Thus, identification of the direct gene targets of ERα in whole organisms will contribute to our understanding of the function of estrogen during development and maturation.
Although transcription factor binding sites can be determined using ChIP on Chip, this technique requires a large set of arrays, whereas our study did not. A recent report described the use of ChIP with a paired-end ditag sequencing strategy (Wei et al. 2006) and this technique may be useful for increasing the numbers of known binding sites and obtaining a genome-wide perspective.
We were unable to estimate the amount of nonspecific contamination of DNA fragments that occurred with the ChIP procedure because it was not possible to map any of the fragments precipitated from the untreated uterus, a result that suggests that our protocol had a very low background. In order to reduce background we introduced a density gradient to remove uncoupled DNA fragments prior to immunoprecipitation. In addition, we introduced steps such as perfusion of the mouse tissue, sonication, and washing of the bound samples, all of which contributed to the reduced background.
Previously, we investigated the temporal regulation of uterine genes by estrogen (Watanabe et al. 2003), but were unable to distinguish whether the activation was a direct effect of ERα or a secondary effect of estrogen. In combination with the results from ChIP, previous DNA microarray data may now be interpreted in greater detail.
Aquaporin in the uterus
The effect of estrogen can be observed clearly in the uterine response. Within 1–2 h of estrogen administration, uterine wet weight begins to increase via water imbibition. Although vascular endothelial growth factors are considered to be crucial to this process (Cullinan-Bove & Koos 1993), the contribution of actual water channels (AQPs) remains poorly understood. However, AQPs are considered to play an important role in the periimplantation stage and it is known that AQP1,4 and 5 are expressed at that stage (Li et al. 1994; Jablonski et al. 2003; Richard et al. 2003; Lindsay & Murphy 2004).
In this study, we demonstrated that ERα regulates AQP5 directly. This result does not concur with the findings of some other studies. For example, using in situ hybridization, Richard et al. (2003) identified AQP1, not AQP5, as the estrogen response gene. This discrepancy may have arisen through differences in the methods of detection, since in general, quantitative PCR is more accurate for the evaluation of changes in mRNA than in situ hybridization. Actually, recent study using mouse cervix detected the expression of AQP5 by Northern hybridization (Anderson et al. 2006). In addition, Jablonski et al. (2003) did not detect expression of AQP5 at the protein level in the uterus; again, these discrepancies may be a result of differences in experimental design. However, our study clearly detected activation of AQP5 via both DNA microarray and quantitative PCR, strongly supporting the suggestion that AQP5 is an estrogen target gene in the uterus. One of the prominent effects of estrogen is water imbibition and for the regulation of water transportation, AQP5 may play critical role.
Estrogen response and aquaporin
As indicated in Table 2, half ERE could be found in 5′ flanking regions of several AQP genes, but only the ERE motif found in AQP5 was functional. Our chromatin immunoprecipitation analysis suggested that the half EREs except for AQP5 were not functional. One possibility is that the EREs were not accessible to ER, but this is less likely because we could detect acetylation of Lys 9 of histon H3 in the region. Although requirement of the cis element necessary for ER binding is still not clear, accumulation and analysis of functional EREs in vivo such as this study may be important. On the other hand, we could not identify functional ERE of AQP8 in its 5′ flanking region. As we could not detect genome DNA fragment adjacent to AQP8, it is possible that other functional ERE exist in other regions or this gene was activated in secondary effect of estrogen. Recently, we found that adrenomedulin (ADM) gene is a direct target gene of estrogen. ADM is activated 1–2 h after estrogen administration and begins to decrease its expression before 6 h. Interestingly, this temporal gene expression profile is very similar to AQP5. On the other hand, the temporal gene expression profile of AQP8 was slightly different and its gene expression level did not change over 12 h. Relevance of direct ER binding and temporal gene expression pattern is one of the intriguing questions.
AQP5 is expressed at high levels in the salivary and lachrymal glands (Raina et al. 1995), which are target tissues of Sjogren's syndrome, an autoimmune disorder that occurs primarily in females. A relationship between AQP5 and this syndrome has been reported (Steinfeld et al. 2001; Tsubota et al. 2001). In addition, a recent study has suggested that estrogen may affect lymphopoiesis (Shim et al. 2004). Therefore, although it remains unclear whether AQP5 expression in the salivary and lachrymal glands is activated by estrogen, the presence of an ERE motif in the promoter region of this gene suggests the possibility of a functional relationship under certain conditions.
Female C57BL/6 J mice were housed under a 12 h light/dark cycle at 23–25 °C, fed laboratory chow (CE-2; CLEA, Tokyo, Japan) and provided with access to tap water. In order to assess the effect of estrogen on uterine gene expression, 8-week-old mice were ovariectomized and two weeks later injected intraperitoneally with 50 µg/kg body weight of 17β-oestradiol (E2) (Sigma-Aldrich Japan, Tokyo, Japan) or sesame oil (Nacalai Tesque, Kyoto, Japan) as a vehicle control. The whole uterus was collected immediately or at the following times after treatment 1, 2, 6, 12 or 24 h. All animal experiments were approved by the institutional Animal Care Committee.
Chromatin-immunoprecipitation-mediated target cloning
The mouse uterus was fixed with 1% formaldehyde and homogenized in phosphate buffered saline (PBS) containing 0.125 m glycine, using a Physcotron (NS-310E, Microtec, Chiba, Japan). The homogenates were centrifuged at 700 g for 5 min at 4 °C and the pellets incubated in lysis buffer (10 mm Tris-HCl [pH 8.0], 10 mm EDTA [pH 8.0], 0.5 mm EGTA and 0.25% Triton X-100) for 10 min. The samples were harvested by microcentrifugation, the pellets resuspended in sonication buffer (10 mm Tris-HCl [pH 8.0], 100 mm NaCl, 1 mm EDTA [pH 8.0], 0.5 mm EGTA), and sonicated using a Bioruptor sonicator (Cosmo Bio, Tokyo, Japan), resulting in DNA with an average length of ∼500 bp. Sonicated samples were centrifuged at 15 000 r.p.m. to remove debris and loaded on to a cesium chloride step gradient containing 1.5 mL of 1.75 g/mL CsCl, 1.5 mL of 1.5 g/mL CsCl, and 1 mL of 1.3 g/mL CsCl supplemented with 0.5% (v/v) N-lauroylsarcosine. Samples were centrifuged at 44 000 r.p.m. in a SW55Ti rotor for 24 h. 0.1 mL fractions were collected from the bottom of the gradient and fractions containing cross-linked chromatin were combined and dialyzed vs. TE buffer. Then the samples were precleared with protein G Sepharose for 1 h at 4 °C and incubated with either 10 µg anti-ERα polyclonal rabbit antibody (H-184; Santa Cruz, CA, USA) overnight at 4 °C. The sample was precipitated with protein G Sepharose, washed 5 times with RIPA buffer (10 mm Tris-HCl [pH 8.0], 140 mm NaCl, 1 mm EDTA [pH 8.0], 0.5 mm EGTA, 1% Triton X-100, 0.1% SDS, 0.1% sodium deoxycholate), then recovered by incubation in elution buffer (0.1 m sodium bicarbonate, 1% SDS). Cross-linking was reversed by incubation for 6 h at 65 °C, followed by incubation with proteinase K for 4 h at 45 °C. Samples were extracted with phenol/chloroform and ethanol precipitated. The precipitated DNA fragments were cloned into pGEM-T vactor (Promega, Tokyo, Japan) by ligation-mediated PCR (LMPCR). Briefly, the DNA fragment was blunt ended and phosphorylated with T4 DNA polymerase (Takara Bio Inc., Shiga, Japan) and T4 polynucleotide kinase (Takara Bio Inc., Shiga, Japan), respectively. After ligating oligonucleotide linkers, the DNA fragment was amplified by PCR and cloned. Oligonucleotide sequences of the linker were F 5′-GCGGTGACCCGGGAGATCTGAATTC-3′ and R 5′-GAATTCAGATC-3′. PCR amplification was performed in the presence of 0.1 nm primer “F,” 0.2 mm of each nucleotide (dATP, dCTP, dGTP and dTTP), 1 × PCR buffer and 1 U of LA Taq (Takara Bio Inc., Shiga, Japan) in 20 µL of reaction buffer. After 30 amplification cycles, the PCR products were purified with QIAquick PCR purification kit (QIAGEN, Tokyo, Japan) and cloned into pGEM-T easy vector. Colonies were randomly picked up after transformation and inserted sequences were analyzed.
Mouse uterine genomic DNA was prepared and sonicated as described above. The sonicated samples were precleared with protein G Sepharose for 1 h at 4 °C, then incubated with either 10 µg anti-ERα polyclonal rabbit antibody (Santa Cruz, CA, USA) or 10 µg anti-acetylated histone H3 (Lys9) antibody (Cell Signaling, Danvers, MA, USA) overnight at 4 °C. The samples were precipitated with protein G Sepharose, washed 5 times with RIPA buffer (10 mm Tris-HCl [pH 8.0], 140 mm NaCl, 1 mm EDTA [pH 8.0], 0.5 mm EGTA, 1% Triton X-100, 0.1% SDS, 0.1% sodium deoxycholate), then recovered by incubation in elution buffer (0.1 m sodium bicarbonate, 1% SDS). Cross-linking was reversed by incubation for 6 h at 65 °C, followed by incubation with proteinase K for 4 h at 45 °C. Samples were extracted with phenol/chloroform and ethanol precipitated. In general, 1/30th of the precipitated DNA was used in a PCR amplification reaction. PCR amplification was performed in the presence of 0.1 nm primers, 0.2 mm of each nucleotide (dATP, dCTP, dGTP and dTTP), 1 × PCR buffer and 1 U of AmpliTaq Gold (PerkinElmer Japan, Tokyo, Japan) in 20 µL of reaction buffer. After 35 amplification cycles, the PCR products were analyzed using agarose electrophoresis. As a negative control, the same experiments were performed with IgG. Primer sequences used to amplify the precipitated DNA are listed in Tables 1 and 2.
Preparation of labeled cRNA and microarray analysis
Total uterine RNA was extracted using TRIzol reagent (Invitrogen, Tokyo, Japan) and the RNeasy total RNA purification kit (Qiagen, Tokyo, Japan). cRNA probes were prepared from purified RNA using a CodeLink Expression assay reagent kit (Amersham Bioscience K.K., Tokyo, Japan). The amplified cRNA (10 µg) was hybridized to oligonucleotide DNA microarrays (CodeLink Uniset Mouse I Bioarray, Amersham Bioscience K.K., Tokyo, Japan), which were scanned and processed using a GenePix 4000B scanner and GenePix Pro software (Axon Instruments, Union City, CA, USA), respectively. In order to confirm the estrogen-related changes in gene expression revealed by DNA microarray analysis, the experiment was repeated independently at least twice. The expression data were analyzed using GeneSpring software (Agilent Tech. Japan, Tokyo, Japan).
Quantitative real time-PCR
cDNA was synthesized from total RNA purified using Superscript II RT(-) (Invitrogen) and random primers at 42 °C for 60 min. Quantitative PCR reactions were performed according to the manufacturer's instructions, using the PE Prism 7000 sequence detector (Applied Biosystems, Tokyo, Japan), SYBR-Green PCR Core Reagents (Applied Biosystems) and primers designed for amplification of short PCR products (< 100 bp; Table 3).
|Gene||RefSeq number||Q-PCR Direction||Position||Sequence||Reporter position||Sequence|
A DNA fragment encoding mouse ERα was amplified from uterine cDNA using the primers 5′-GGCGAATTCATGACCATGACCCTTCACAC-3′ and 5′-GCAGTCGACTCAGATCGTGTTGGGGAAGC-3′, digested with EcoRI and SalI, then ligated into pcDNA3.1(+) (Invitrogen), generating pERαcDNA3.1.
The 5′ flanking regions of AQP genes were amplified using the primers listed in Table 3, then digested with MluI and BglII and ligated into pGL2-basic (Promega, Tokyo, Japan). Inserts for the shorter AQP5 reporters, p103AQP5-luc and p13AQP5-luc (5′ regions flanking AQP5, -103 to +74 and -13 to +74, respectively), were amplified using the forward primers 5′-cgacgcgtTGGGTGAGACCGACCGGGTCAAGATG-3′ and 5′-cgacgcgtAAAGGCCGGCCGGAGAGGGA-3′, respectively. The primer 5′-cgacgcgtTGGGTGAGACCGACCGaaTCAAGATGCTCC-3′ was used for amplification of the mutated ERE reporter (pm103AQP5-luc).
Luciferase reporter assay
HEK293T cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% fetal bovine serum. For ligand-dependent transcription assays, cells were seeded in 24-well tissue culture plates and grown in phenol red-free DMEM supplemented with 5% charcoal/dextran-treated fetal calf serum (JRH Biosciences, Inc. Lenexa, KS, USA). HEK293T cells were transfected using FuGene6 (Roche Diagnostics K.K., Tokyo, Japan), following the manufacturer's instructions, and the following day treated with either ethanol or 10 nm E2. Typically, cells were transfected with 0.3 µg luciferase reporter plasmid, 50 ng pERαcDNA3.1 (ERα expression plasmid) or 50 ng pRL-TK reporter plasmid (cDNA encoding Renilla luciferase downstream of the thymidine kinase promoter) (Promega, Tokyo, Japan). Luciferase values were normalized to the Renilla luciferase activity.
This work was supported in part by a Core Research for Evolutional Science research grant from the Japan Science and Technology Cooperation, a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science, a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan, a Health Sciences Research Grant from the Ministry of Health, Labour and Welfare of Japan, a research grant from the Ministry of Environment, Japan.
- 2006) Utilization of different aquaporin water channels in the mouse cervix during pregnancy and parturition and in models of preterm and delayed cervical ripening. Endocrinology 147, 130–140. , , & (
- 2004) Genome-wide identification of high-affinity estrogen response elements in human and mouse. Mol. Endocrinol. 18, 1411–1427. , , , , , , , & (
- 2005) Chromosome-wide mapping of estrogen receptor binding reveals long-range regulation requiring the forkhead protein FoxA1. Cell 122, 33–43. , , , , , , , , , , , & (
- 1993) Vascular endothelial growth factor/vascular permeability factor expression in the rat uterus: rapid stimulation by estrogen correlates with estrogen-induced increases in uterine capillary permeability and growth. Endocrinology 133, 829–837. & (
- 1999) Expanding the TRANSFAC database towards an expert system of regulatory molecular mechanisms. Nucleic Acids Res. 27, 318–322. , , , , , , , , & (
- 1999) Identifying DNA and protein patterns with statistically significant alignments of multiple sequences. Bioinformatics 15, 563–577. & (
- 2003) Estrogen receptor-dependent genomic responses in the uterus mirror the biphasic physiological response to estrogen. Mol. Endocrinol. 17, 2070–2083. , , , , , & (
- 1993) Genomic binding-site cloning reveals an estrogen-responsive gene that encodes a RING finger protein. Proc. Natl. Acad. Sci. USA 90, 11117–11121. , , , , , , , , & (
- 2003) Estrogen regulation of aquaporins in the mouse uterus: potential roles in uterine water movement. Biol. Reprod. 69, 1481–1487. , , & (
- 2000) Estrogen receptor α rapidly activates the IGF-1 receptor pathway. J. Biol. Chem. 275, 18447–18453. , , , , & (
- 2005) Identification of estrogen-responsive genes using a genome-wide analysis of promoter elements for transcription factor binding sites. J. Biol. Chem. 280, 21491–21497. , & (
- 1998) Generation and reproductive phenotypes of mice lacking estrogen receptor beta. Proc. Natl. Acad. Sci. USA 95, 15677–15682. , , , , , , , , & (
- 2005) From the Cover: Location analysis of estrogen receptor alpha target promoters reveals that FOXA1 defines a domain of the estrogen response. Proc. Natl. Acad. Sci. USA 102, 11651–11656. , , , , & (
- 1994) The water channel gene in human uterus. Biochem. Mol. Biol. Int. 32, 371–377. , & (
- 2004) Redistribution of aquaporins in uterine epithelial cells at the time of implantation in the rat. Acta Histochem. 106, 299–307. & (
- 1993) Alteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor gene. Proc. Natl. Acad. Sci. USA 90, 11162–11166. , , , , & (
- 1996) Tyrosine kinase/p21ras/MAP-kinase pathway activation by estradiol-receptor complex in MCF-7 cells. EMBO J. 15, 1292–1300. , , , , , & (
- 2004) Phenotypic anchoring of gene expression changes during estrogen-induced uterine growth. Environ. Health Perspect. 112, 1589–1606. , , , , , , , , , , , , & (
- 2004) Genomic targets of nuclear estrogen receptors. Mol. Endocrinol. 18, 1859–1875. , , & (
- 2000) mRNAs encoding aquaporins are present during murine preimplantation development. Mol. Reprod. Dev. 57, 323–330. , , , , & (
- 1995) Molecular cloning and characterization of an aquaporin cDNA from salivary, lacrimal, and respiratory tissues. J. Biol. Chem. 270, 1908–1912. , , & (
- 2003) Aquaporin water channel genes are differentially expressed and regulated by ovarian steroids during the periimplantation period in the mouse. Endocrinology 144, 1533–1541. , , & (
- 2002) Epigenetic codes for heterochromatin formation and silencing: rounding up the usual suspects. Cell 108, 489–500. & (
- 2004) Aromatase-deficient mice spontaneously develop a lymphoproliferative autoimmune disease resembling Sjogren's syndrome. Proc. Natl. Acad. Sci. USA 101, 12628–12633. , , , , , , , , & (
- 2000) Interaction of oestrogen receptor with the regulatory subunit of phosphatidylinositol-3-OH kinase. Nature 407, 538–541. , , , , & (
- 2001) Abnormal distribution of aquaporin-5 water channel protein in salivary glands from Sjogren's syndrome patients. Lab. Invest. 81, 143–148. , , , , & (
- 2001) Defective cellular trafficking of lacrimal gland aquaporin-5 in Sjogren's syndrome. Lancet 357, 688–689. , , , & (
- 2004) Tissue-specific estrogenic and non-estrogenic effects of a xenoestrogen, nonylphenol. J. Mol. Endocrinol. 33, 243–252. , , , , & (
- 2003) Analysis of temporal changes in the expression of estrogen-regulated genes in the uterus. J. Mol. Endocrinol. 30, 347–358. , , , , , , & (
- 2002) Genome-wide analysis of changes in early gene expression induced by oestrogen. Genes Cells 7, 497–507. , , , , , & (
- 2006) A global map of p53 transcription-factor binding sites in the human genome. Cell 124, 207–219. , , , et al. (
- 2002) Estrogen receptor-interacting protein that modulates its nongenomic activity-crosstalk with Src/Erk phosphorylation cascade. Proc. Natl. Acad. Sci. USA 99, 14783–14788. , , , & (