THE JOURNAL • RESEARCH • www.fasebj.org

Estrogen receptor a is required for oviductal transport of embryos † † ‡ § Shuai Li,* Sofia R. S. O’Neill,* Yong Zhang, Michael J. Holtzman, Ken-Ichi Takemaru, Kenneth S. Korach, and Wipawee Winuthayanon*,1 † *School of Molecular Biosciences, College of Veterinary Medicine, Washington State University, Pullman, Washington, USA; Pulmonary and ‡ Critical Care Medicine, Department of Medicine, Washington University School of Medicine, St. Louis, Missouri, USA; Department of Pharmacological Sciences, Stony Brook University School of Medicine, Stony Brook, New York, USA; and §Reproductive and Developmental Biology Laboratory, National Institute of Environmental Health Sciences, National Institutes of Health, Research Triangle Park, North Carolina, USA

ABSTRACT: Newly fertilized embryos spend the first few days within the oviduct and are transported to the uterus, where they implant onto the uterine wall. An implantation of the embryo before reaching the uterus could result in ectopic pregnancy and lead to maternal death. Estrogen is necessary for embryo transport in mammals; however, the mechanism involved in estrogen-mediated cellular function within the oviduct remains unclear. In this study, we show in mouse models that ciliary length and beat frequency of the oviductal epithelial cells are regulated through estrogen receptor a (ESR1) but not estrogen receptor b (ESR2). profiling indicated that transcripts in the WNT/ b- (WNT/CTNNB1) signaling pathwaywereregulatedbyestrogeninmouse oviduct, and inhibition of this pathway in a whole oviduct culture system resulted in a decreased embryo transport distance. However, selective ablation of CTNNB1 from the oviductal ciliated cells did not affect embryo transport, possibly because of a com- pensatory mechanism via intact CTNNB1 in the adjacent secretory cells. In summary, we demonstrated that disruption of estrogen signaling in oviductal epithelial cells alters ciliary function and impairs embryo transport. Therefore, our findings may provide a better understanding of etiology of the ectopic pregnancy that is associated with alteration of estrogen signals.—Li, S., O’Neill, S. R. S., Zhang, Y., Holtzman, M. J., Takemaru, K.-I., Korach, K. S., Winuthayanon, W. Estrogen receptor a is required for oviductal transport of embryos. FASEB J. 31, 1595–1607 (2017). www.fasebj.org

KEY WORDS: cilia • epithelial cells • fertility

The oviducts (or fallopian tubes in humans) are divided After ovulation, the ovulated oocytes are drawn into into 3 segments: infundibulum (adjacent to the ovary), the oviduct by the current generated by ciliated cells at the ampulla (site of fertilization), and isthmus (connected to infundibulum (3). The sperm must transit from the uterus the uterus). Each of these segments is made up of epithelial, through the length of the oviduct to the site of fertilization stromal, and smooth muscle cells. In mammals, the epi- in the ampulla. The resulting embryos are subsequently thelial lining of the oviduct consists of secretory and transported from the ampulla, through the isthmus, and multiciliated cells. The infundibulum and ampulla are into the uterus, where implantation takes place. The rela- composed of more ciliated than secretory cells, whereas tive importance of muscular contractility and cilia motility the isthmus has more secretory than ciliated cells (1, 2). The varies along the length of the oviduct. For example, in pigs, multiciliated cells possess motile cilia with a 9+2 arrange- isthmus contractions proceed toward and tend to fade ment of , composed of a-andb-. beyond the ampulla-isthmic junction (AIJ) (4). In rats and rabbits, inhibition of smooth muscle activity has no effect on embryo transport (5–7), indicating that ciliary motility a a – ABBREVIATIONS: Ac -, acetylated -tubulin; AIJ, ampulla isthmic is sufficient for transporting the embryos without muscle junction; CTNNB1, b-catenin; CBF, ciliary beat frequency; dpc, days postcoitus; ESR, estrogen receptor; IHC, immunohistochemical; IWP2, contractions. inhibitor of WNT production 2; KSOM, potassium simplex optimization Ciliary dysfunction can lead to human infertility or medium; PCD, primary ciliary dyskinesia; PGR, progesterone receptor; ectopic pregnancy. A rare genetic disorder called primary SERM, selective estrogen receptor modulator; SFRP, secreted frizzled- related ; TBST, Tris-buffered saline-Tween 20; UTJ, uterotubal ciliary dyskinesia (PCD) causes abnormalities in ciliary junction; WNT, wingless-type MMTV integration site family motility in the ear canal, respiratory tract, and re- 1 Correspondence: School of Molecular Biosciences, Biotechnology/Life productive tract. In men, ciliary dysfunction affects sperm Science BLS202, Washington State University, Pullman, WA 99164, USA. motility (8) whereas, it affects the functionality of the fal- E-mail: [email protected] lopian tube in women (9–11). Some women with PCD are doi: 10.1096/fj.201601128R infertile (9, 11–13), partly due to fewer cilia per cell and

0892-6638/17/0031-1595 © FASEB 1595 Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 decreased ciliary beat frequency (CBF) (9, 11, 12). In ad- eggs, were excluded, because we could not distinguish between dition, ectopic pregnancy, a condition where the embryo the previously ovulated eggs and superovulated eggs that were implants outside the uterine wall, has been associated with not fertilized. defective ciliary function of the fallopian tubes (14). For embryo transfer, 1-cell-stage embryos (zygotes) were collected from CF-1 donor mice and transferred into the oviduct Moreover, ectopic pregnancy, because of internal bleeding f/f Cre/+ f/f of Esr1 and Wnt7a ;Esr1 pseudopregnant females that caused by rupture of the implantation site, is one of the were mated with vasectomized males (22). The embryos were causes of maternal death during the first trimester. These collected at 3.5 dpc from the oviduct and uterus, and only viable conditions suggest that the presence of functional ciliated embryos were included in the experiment. Nonviable embryos cells in the fallopian tubes is necessary for normal embryo were excluded from the experiment, because they could not be transport and successful establishment of pregnancy. distinguished accurately from ovulated, unfertilized eggs gen- The cell morphology of the oviductal epithelium is af- erated by the pseudopregnant recipients. fected by dynamic changes in the ovarian hormones, es- trogen (E2) and progesterone (P4), during the estrous/ Immunohistochemical and immunofluorescent analysis menstrual cycle in mammals (15, 16). E2 and P4 act through their nuclear receptors [estrogen receptor a (ESR1) and estrogen receptor b (ESR2), and progesterone receptor Another set of animals was bred with wild-type males, and the (PGR), respectively]. ESR1 and PGR are expressed in all embryos were left intact without flushing for histologic analysis. – The oviducts were collected at 0.5 or 3.5 dpc and fixed in 10% cell layers of the oviduct (17 19). ESR2 is exclusively de- buffered formalin solution. The uterus was collected from Esr2+/+ 2 2 tected in the cilia of ciliated epithelial cells (20). Previous and Esr2 / females at 3.5 dpc. Then, the tissues were embed- reports showed that E2 increases the number of ciliated dedinparaffinandcutat5mm. The tissue sections were cells in the mouse neonatal oviduct (17). However, the deparaffinized and incubated with citrate buffer (pH 6.0), by requirement of ESR1 and -2 for the proper ciliary function using the heat-induced epitope-retrieval method. The tissues during embryo transport remains unclear. Moreover, the were blocked with normal horse serum and incubated with pri- mary antibodies against acetylated a-tubulin (Ac a-tubulin, 1: molecular mechanisms regulated by E2 through ESR1 and 6000, T6793; Sigma-Aldrich, St. Louis, MO, USA) and CTNNB1 -2 in the ciliated epithelial cells during embryo transport (1:200, 610154; BD Biosciences, San Jose, CA, USA). The detection have not yet been explored. In this study, we used global of ESR1, ESR2, and Ki67 was performed as previously reported and conditional Esr-knockout mouse models to address (25). DAB chromogen was used as an antigen labeling system the functionality of ESR1 and -2, as well as the possible (Dako, Carpentaria, CA, USA). The tissues were mounted and downstream signaling cascades that regulate the function coverslipped. Hematoxylin and eosin (H&E) staining was per- of ciliated cells in the oviduct during embryo transport. formed according to standard histologic procedure. The images were obtained with the DMi8 microscope (Leica, Bannockburn, IL, USA). In some experiments, expression of Ac-a-tubulin was MATERIALS AND METHODS evaluated with donkey anti-mouse 488 (Thermo Fisher Scientific, Waltham, MA, USA) as a secondary antibody, and the tissues were coverslipped with Prolong Gold with DAPI (Thermo Fisher Mouse models Scientific). The fluorescent images were taken with a UV 510 microscope (Zeiss, Thornwood, NY, USA). We generated the conditional deletion of Esr1 in the oviductal and uterine epithelial cells using Wnt7aCre/+;Esr1f/f,aspreviously published (21, 22). Esr1f/f animals were used as control litter- Quantification of Ki67-positive cells 2 2 2 2 mates. Esr2+/+ and Esr2 / (or ERb / ) animals (23) were pur- chased from Taconic (Germantown, NY, USA). Ctnnb1f/f animals The oviduct tissues stained with Ki67 antibody were imaged at were obtained from The Jackson Laboratory (Bar Harbor, ME, 3100 magnification. Individual images were imported into Fiji USA). Ciliated cell-selective cre-expressing animals (or Foxj1Cre+) software (26). The numbers of epithelial cells and Ki67-positive (24) were crossed with the Ctnnb1f/f animals to generate the cells were counted using the “Cell-counter” plug in. The Ki67- conditional deletion of Ctnnb1 (encoding CTNNB1 protein) in the positive cells were calculated and presented as a percentage of ciliated cells. positive cells per total epithelial cells in that microscopic field. At least 5 areas (images) were used from each animal for cell counting (n=4–5 animals per group). Embryo collection and embryo transfer

2 2 Adult female mice (Esr1f/f, Wnt7aCre/+;Esr1f/f, Esr2+/+, Esr2 / , Ciliary length measurement Ctnnb1f/f,andFoxj1Cre+;Ctnnb1f/f)attheageof8–12 wk were superovulated with gonadotropins [5 U of pregnant mare’sse- Formalin-fixed oviductal sections from 4 animals per genotype rum gonadotropin and 5 U of human chorionic gonadotropin were stained with Ac a-tubulin, as described above. Images were (hCG); EMD–Millipore, Billerica, MA, USA] (22). Immediately taken at the infundibulum and ampullary regions with a Leica after hCG injections, the females were housed with wild-type DMi8 microscope with a 3100 objective lens. The scale bars male breeders overnight, and the copulatory plugs were checked (10 mm) were automatically embedded into all images for further the following morning and designated as 0.5 d postcoitus (dpc). ciliary length analysis in Fiji software. A total of 51, 82, 87, and 96 f/f f/f At 3.5 dpc (10 AM), the female reproductive tracts were dissected images were obtained of the Esr1 infundibulum, Esr1 am- from the animals. The embryos were first flushed from the uteri, pulla, Wnt7aCre/+;Esr1f/f infundibulum, and Wnt7aCre/+;Esr1f/f then the oviducts. The number of embryos from each region ampulla, respectively. The observer was blinded to the animal’s (uterus vs. oviduct) was calculated as a percentage of the total genotype to minimize bias. To calibrate the length of 10-mmscale morula and blastocyst stage embryos collected. Only the viable bars, the “straight-line” tool was used to draw a line and set to a embryos were included in the experiment. Nonviable embryos, known distance of 10 mm using the “Set Scale” tool. Then, the such as empty zona pellucida, unfertilized eggs, and fragmented ciliary length of each ciliated epithelial cell was measured

1596 Vol. 31 April 2017 The FASEB Journal x www.fasebj.org LI ET AL. Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 from the apical membrane (base of the cilia) to the tip of the Immunoblot analysis cilia using the “Straight-line” tool. Using all images, 674, 1446, 1098, and 1540 ciliated epithelial cells were measured for cil- Another set of animals was bred with wild-type males (3 f/f f/f iary length analysis from Esr1 infundibulum, Esr1 am- animals/genotype). The whole oviducts were collected at 0.5 Cre/+ f/f Cre/+ pulla, Wnt7a ;Esr1 infundibulum, and Wnt7a ; dpc, snap frozen, and stored at 280°C until use. The protein was f/f Esr1 ampulla, respectively. extracted from the oviducts with tissue protein extraction reagent (Thermo Fisher Scientific) with Halt protease and phosphatase m Microarray analysis inhibitor cocktail. A total of 2 g protein/lane was loaded into 10% SDS-polyacrylamide gels. The protein was transferred onto f/f Cre/+ f/f nitrocellulose membranes using the Trans-Blot Turbo Transfer The oviducts were collected from Esr1 and Wnt7a ;Esr1 System (Bio-Rad Laboratories, Hercules, CA, USA). The mem- females at 0.5 dpc (n = 4 per genotype). RNA extraction and branes were blocked with 5% nonfat dry milk in Tris-buffered microarray analysis were previously described (22). The data saline with Tween 20 (5% milk TBST) and incubated at 4°C were deposited in the National Center for Biotechnology overnight with the following primary antibodies in 5% milk Information (Bethesda, MD, USA) Gene Expression Omnibus TBST: Ac a-tubulin (TUBA1B, 1:1000, T6793; Sigma-Aldrich), (Accession No. GSE37471). In brief, the microarray dataset was ADP-ribosylation factor-like protein 13B (ARL13B, 1:1000, 17711- analyzed using Partek Genomic Suite software (Partek, St. Louis, 1-AP, Proteintech), b-tubulin IV (TUBB4, 1:1000, MU178-UC, MO, USA). that were expressed $ |1.5|-fold in Wnt7aCre/+; b f/f f/f BioGenex, Fremont, CA, USA), and - (1:5000, SC-47778, Esr1 compared to Esr1 oviducts were considered signif- Santa Cruz Biotechnology, Dallas, TX, USA). Then, the mem- icantly different. Differentially expressed genes were ana- branes were washed 3 times for 10 min each in TBST and lyzed in Ingenuity Pathway Analysis (Qiagen, Valencia, CA, incubated with horseradish-peroxidase conjugated goat anti- USA). mouse (Santa Cruz Biotechnology) or goat anti-rabbit (Abcam, Cambridge, MA, USA) diluted at 1:5000 in 5% milk TBST at room temperature for 1 h and washed 3 times for 10 min each in TBST. Oviduct collection for gene expression analysis Chemiluminescence detection was performed with ECL clarity (Bio-Rad) and detected with ChemiDoc MP System (Bio-Rad). Adult female mice (8–12 wk old) were used for the gene ex- The band intensities were quantified and normalized to b-actin pression analysis of WNT/CTNNB1 signaling pathway. The using Fiji software with Analyze and Gel Tools. estrous cycle was determined by vaginal cytology, as previously indicated (27). The oviducts were collected at each stage of es- trous cycle, and all regions of the oviduct (infundibulum, am- Human fallopian tube tissues pulla, and isthmus) were dissected and collected separately. RNA was extracted from the dissected tissues using Trizol re- Human fallopian tubes were surgically removed from 3 women agent (Thermo Fisher Scientific), according to the manufacturer’s who underwent gynecologic procedures (such as tubal ligation, protocol. RNA from each sample (1 mg) was incubated with salpingohysterectomy, or oosalpingohysterectomy) at the Stony DNaseI to eliminate the DNA in the samples, and the samples Brook University School of Medicine. The tissues were collected were reverse transcribed to cDNA with SuperScriptII (Thermo as biopsies or discarded tissues, fixed with formalin, and em- Fisher Scientific). Then, the cDNAs were used as a template for bedded in paraffin. The paraffin-embedded tissue blocks were in real-time PCR analysis. Quantitative PCR was performed with the archival collections of the Stony Brook University’sPathology Fast SBYR Green master mix (Thermo Fisher Scientific) according Department and were deidentified and shipped to Washington to the manufacturer’s protocol. The expression values were State University. The ethics approvals for this study were normalized to ribosomal protein L7 (Rpl7) and calculated as obtained from both Washington State University and Stony previously described (21). The primer sequences are (F, forward; Brook University Medical Center. The expression analysis of R, reverse): Ctnnb1 (F: 59-TCCGAGGACTCAATACCATTCC-39 CTNNB1 and Ac a-tubulin was performed with immunohisto- and R: 59-AGGACCCCTGCAGCTACTCTT-39); Cby1 (F: 59- chemical analysis as indicated above. AGACAGGAGACGGAAGTGCT-39 and R: 59-GCTGAATAT- GCTGCCAAAGA-39); Sfrp2 (F: 59-TCCTGCCGGCCACAGA-39 Ex vivo and R: 59-GTTGTCGTCCTCATTCTTGGTTT-39), Axin2 (F: oviduct culture 59-GTCACAGGATGTCTGGCAGT-39 and R: 59-TTATGCTT- TGGGCACTATGG-39); Cdh1 (F: 59-CTGGGCAGAGTGAG- The procedures for ex vivo culture of the mouse oviduct was ATTTGA-39 and R: 59-CCACTTTGAATCGGGAGTCT-39); adapted from a previous report (28). In brief, the oviducts were Fgf18 (F: 59-TTCTACTGCTGTGCTTCCAGGTT-39 and R: dissected from female mice at 10–11 AM of 0.5 dpc. Then the 59-ACGTGGATGCGGAAGTCC-39); Foxj1 (F: 59-CTCTGAGC- oviducts were washed in Leibovitz-15 (Thermo Fisher Scientific) CAGGCCTCACATT-39 andR:59-GGTCAACATCCA- with 1% fetal bovine serum (Gemini Bio-Products, West Sacra- CAGGCTGA-39); Fzd9 (F: 59-GGGATGGCCAGCTCACTTT-39 mento, CA, USA) and trimmed of gonadal fat. The mesenteric and R: 59-CCTCGTGGCCCCATTTTT-39); Fzd10 (F: 59-GAA- ligaments were carefully cut with fine-point dissecting forceps GCCAACAGCAGCTACTTTCA-39 and R: 59-GCCACCCT- and detached from the oviduct to linearize (uncoil) the oviduct GCGCATCA-39); Kif3a (F: 59-GACATCAGTGGGTCAGAGGA-39 under a heated stage stereomicroscope (Leica Microsystems, Inc., and R: 59-TTTACCTGCTTGGTCCCTTC-39); Kif6 (F: 59-GGAC- Buffalo Grove, IL, USA) (29). Uncoiling the oviducts at the start of TCGGAAAGGTCAAGGG-39 and R: 59-AGTAAATCCCTTGC- culture allowed us to accurately locate the embryos at the end of TGCTGCT-39); Kif9 (F: 59-CCGGTGGCGTTAGCTAAAAT-39 the experiment without additional oviduct manipulations that and R: 59-CCCTCTCCGAGTGTCCTTCT-39); Kif19 (F: 59-TCT- could affect embryo location. The uncoiled oviducts with em- ATGTTCGGACCCTCAATGA-39 andR:59-ATCACCCCTTT- bryos inside were placed on a Transwell insert (Corning, Inc., GGAGTCTTCC-39); Tcf3 (F: 59-GCTGTCACAGACACAGTCTCAGC-39 Corning, NY, USA) in a 24-well plate with potassium simplex and R: 59-TGAATGCATGGCTTCTTGCT-39); Wnt2 (F: optimization medium (KSOM), which is suitable for embryo 9 9 9 5 -AACACCCTGGACAGAGATCACA-3 and R: 5 -GCGTAA- development, at 37°C with 5% CO2. The oviduct from the left side ACAAAGGCCGATTC-39); Wnt7a (F: 59-CTTTCTCAGCC- was cultured with 0.1% DMSO (Sigma-Aldrich) in KSOM and TGGGCATAGTCT-39A and R: ATGATGCTCGCACCCAGAG-39); therightsideoftheoviductfromthesameanimalwascul- and Wnt7b (F: 59-AACCTTCACAACAATGAGGCG-39 and R: tured with 5 mM inhibitor of WNT production 2 (IWP2; Tocris, 59-CACACCGTGACACTTACATTCCA-39). Bristol, United Kingdom) diluted in 0.1% DMSO in KSOM.

ESTROGEN RECEPTOR-a IN TRANSPORT OF EMBRYOS 1597 Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 The medium containing DMSO or IWP2 was refreshed every genotype. The Mann-Whitney U test was used to separate 24 h. After 3 d, the oviducts were placed on a slide, and the medians for the CBF data. Comparison of gene expressions location of the embryos was determined. Embryo transport during estrous cycle was made using 1-way ANOVA, Tukey was calculated with a dissecting microscope equipped with an post hoc test. eyepiece micrometer, by measuring from the AIJ to the far- thest embryo in the oviduct toward the uterotubal junction (UTJ). The distance of the farthest embryo was used because most embryos travel singly, not as a group, within the ovi- RESULTS ductal lumen. After the location of the embryos was measured in the oviducts, the embryos were flushed to evaluate the ef- ESR2 localizes to cilia in the oviduct fect of the treatment on embryo development. The oviducts epithelium but is dispensable for the were also collected for RNA expression analysis of Axin2.The embryo transport function of the oviduct experiments were repeated twice from a total of 4–5animals per treatment. To first confirm whether ESR1 and -2 are expressed in the ciliated epithelial cells in mice, we assessed the lo- calization of ESR1 and -2 in comparison with Measurement of CBF Ac a-tubulin, a cilia marker within the oviduct. We observed a positive staining of ESR1 in the cell body of To evaluate the CBF in Esr1f/f and Wnt7aCre/+;Esr1f/f females, we measured the beat frequency in the ex vivo oviducts collected at all oviductal cell layers, including epithelial, stromal, 0.5 dpc (n=3 per genotype). The oviducts were opened and muscle cells, whereas ESR2 was only detected in Ac longitudinally. The infundibulum and isthmus were mounted a-tubulin-positive ciliary stalks of ciliated epithelial (leaving the ampulla unmounted), with tissue adhesive onto a cells (at the infundibulum and ampulla region; Fig. 1A). glass-bottom dish. During imaging, the oviduct tissue was cul- Based on the high abundance of ESR2 protein specifi- tured in 4 ml Leibovitz-15 at 37°C with a heated chamber. Images cally in the cilia of oviductal ciliated cells, we hypothesized were recorded from the ampulla at 100 frames per second for 10 s that ESR2 has a role in the formation and function of cilia with a high-speed camera (Andor Zyla, Andor, Concord, MA, USA) with a 3100 objective connected to an inverted bright-field and is therefore critical for the embryo transport in the microscope (DMi8, Leica). The regions of interest (ROIs) at 434 oviduct. To investigate the functional requirement of ESR2 pixels were selected from the sample at the tip of the cilia. The during the embryo transport, we determined the number light signal intensities from each ROI over time were collected of embryos present in the uterus, the oviduct, or both at 3.5 2 2 using Fiji software and transformed into a frequency using a fast dpc of Esr2 / mice. Although the number of embryos Fourier transform algorithm in AutoSignal software (Systat 2/2 – was slightly lower in Esr2 because of an ovulatory Software Inc., San Jose, CA, USA). The total of 168 469 ROIs were defect (30), there was no significant difference between the recorded from each animal. 2 2 number of morulae and blastocysts in Esr2 / when compared to Esr2+/+ females (Fig. 1B). Almost 100% of the embryos were retrieved from the uterus at 3.5 dpc from Statistical analysis 2 2 the Esr2+/+ and Esr2 / animals (Fig. 1C). Uterine cross- Results reached statistical significance at P , 0.05 using an sections indicated that the embryos were present in +/+ 2/2 unpaired Student’s t test (GraphPad Prism, La Jolla, CA, the uterine lumen in both Esr2 and Esr2 animals at USA) for comparison of 2 separated groups of treatment or 3.5 dpc (Fig. 1D). These findings indicate that ESR2 is

Figure 1. ESR2 is exclusively present in the cilia of oviductal epithelial cells but is not necessary for embryo transport. A) Immunofluorescent staining of Ac a-tubulin and IHC analysis of ESR1 and -2 of wild-type mouse oviduct in the infundibulum, ampulla, and isthmus. Insets are negative controls. Note that multiciliated cells (indicated by Ac a-tubulin IHC) were more abundant in the infundibulum and ampulla compared to isthmus. B) Number of embryos at morulae and blastocyst stages 2 2 collected from the oviduct and the uterus of control, Esr2+/+ (n=3 mice), and Esr2 / (n=4 mice) females at 3.5 dpc. C) +/+ 2/2 Percentage of the embryos retrieved from the uterus in Esr2 and Esr2 females at 3.5 dpc. Graphs represent means 6 SEM. 2 2 D) H&E staining of embryos (E) within the uterine lumens of Esr2+/+ and Esr2 / females at 3.5 dpc.

1598 Vol. 31 April 2017 The FASEB Journal x www.fasebj.org LI ET AL. Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 dispensable for the early development and transport of ciliary length is crucial for cilia function in the oviduct (34). embryos. We found that the ciliary length at the infundibulum, mea- sured from the oviductal tissues stained with Ac a-tubulin, Epithelial ESR1 is necessary for the embryo was significantly longer in the Wnt7aCre/+;Esr1f/f compared transport in the oviduct with the Esr1f/f oviducts (Fig. 2I). The ciliary length in the ampulla region had a tendency to be longer in Wnt7aCre/+; f/f Because it has been established that E2 is necessary for the Esr1 , but it was not significantly different from the length transport of embryos in murine models (31) and we ob- in Esr1f/f oviducts. These findings suggest that selective served no effect of Esr2 ablation on the embryo transport, ablation of Esr1 in the oviductal epithelial cells does not we hypothesized that the E2-dependent embryo transport affect epithelial cell proliferation but alters the length of the occurs through ESR1, specifically epithelial ESR1. Un- cilia in the oviduct. 2 2 fortunately, the global Esr1 / females are infertile and A recent study showed that a global deletion of a the hormonal profile (elevated testosterone, E2 and lutei- kinesin superfamily (Kif) gene caused an increase in ciliary 2 2 nizing hormone) is abnormal. Therefore, the Esr1 / length, resulted in an ineffective clearance of the mucus model is not suitable as a tool to study embryo transport and cell debris in the oviduct (34). Therefore, to assess Cre/+ f/f under the influence of E2. To circumvent this, we have whether the embryo transport defect in Wnt7a ;Esr1 recently generated a conditional-knockout mouse model animals is possibly caused by a deficit in ciliary function, where Esr1 is selectively ablated only in the epithelial cells we measured CBF in ex vivo cultured oviducts at 0.5 dpc. A (Wnt7aCre/+;Esr1f/f animals) of the female reproductive box-and-whisker plot of the CBFs showed a significant tract, including the oviduct and uterus (21, 22). Moreover, difference in the median values of the beat frequencies these females showed normal features of estrous cyclicity between Esr1f/f and Wnt7aCre/+;Esr1f/f oviducts (Fig. 2J). and ovulatory response (21). We confirmed that ESR1 In addition, the number of cells at each frequency was protein was not present in the epithelial cells of the oviduct presented in a histogram, which showed that there were in Wnt7aCre/+;Esr1f/f mice but was detected in Esr1f/f mice more ciliated cells in Wnt7aCre/+;Esr1f/f that beat at lower (control littermates) (Fig. 2A). We also previously reported frequencies, compared with those in Esr1f/f animals. that embryos in Wnt7aCre/+;Esr1f/f females die before Moreover, we evaluated whether an increase in the ciliary the 2-cell stage (22). Therefore, to assess whether the em- length in Wnt7aCre/+;Esr1f/f oviducts was due to an aber- bryo transport function of the oviduct was dysfunctional, rant expression of certain kinesin family genes. Kif9 and we transferred 1-cell stage embryos (zygotes) from wild- Kif19,butnotKif3a or Kif6, were expressed at a signifi- type donors into the oviducts of pseudopregnant Esr1f/f cantly lower level in Wnt7aCre/+;Esr1f/f compared to Esr1f/f or Wnt7aCre/+;Esr1f/f animals and retrieved the embryos oviducts (Fig. 2K). In addition, Foxj1 (ciliogenesis marker) at 3.5 dpc. The majority of the embryos (82%) retrieved level was also significantly less in Wnt7aCre/+;Esr1f/f from Esr1f/f females were located in the uterus, whereas compared to Esr1f/f oviducts. These results suggest that a 100% of the embryos retrieved from Wnt7aCre/+;Esr1f/f lack of ESR1 in the oviductal epithelial cells results in an females were retained only in the oviducts—none in the increase in the ciliary length, which alters ciliary beat uterus (Fig. 2B). These results suggest that E2-dependent function; thus, such a condition may be an explanation for embryo transport function in the oviduct is mediated the cause of defective embryo transport. through the epithelial ESR1. E2-induced proliferation of epithelial cells in the female The WNT/CTNNB1 signaling pathway in the reproductive tract is mediated by ESR1 (21). To evaluate oviduct is regulated by E2 whether a lack of epithelial Esr1 causes aberrant epithelial Cre/+ f/f cell proliferation in Wnt7a ;Esr1 females and con- To identify downstream molecular mechanisms of E2 sig- tributes to defective embryo transport, we examined the nals regulating cilia function, we evaluated the potential proliferative index of these cells using Ki67 IHC analysis in signaling pathways that are shown to be involved in cilia Esr1f/f and Wnt7aCre/+;Esr1f/f oviducts at 0.5 dpc (Fig. 2C). function, one of which is the WNT/CTNNB1 pathway There was no indication of proliferation differences in the (35). WNT family members are secreted glycolipoproteins Wnt7aCre/+;Esr1f/f oviducts, because a comparable num- that signal through both canonical CTNNB1 and non- ber of Ki67-positive epithelial cells was observed in Esr1f/f canonical pathways. CTNNB1, a transcriptional coac- and Wnt7aCre/+;Esr1f/f oviducts (Fig. 2D). To next de- tivator encoded by the Ctnnb1 gene, regulates embryonic termine whether the loss of epithelial ESR1 impacts ciliary development and adult homeostasis (36). Recent studies in morphology, we evaluated the presence of cilia in the zebrafish and mice demonstrated that WNT/CTNNB1 oviduct using H&E staining with high-magnification im- signaling is crucial for ciliogenesis and airway multiciliated aging (Fig. 2E)andthepresenceofAca-tubulin (Fig. 2F). cell differentiation (35). Thus, we sought to test whether a We found that the positive staining of Ac a-tubulin transport defect in the Wnt7aCre/+;Esr1f/f oviduct was was comparable between Esr1f/f and Wnt7aCre/+;Esr1f/f due to altered WNT/CTNNB1 signaling in the oviduct. oviducts (Fig. 2F). Moreover, the levels of ciliary proteins Microarray analysis of the oviducts collected from Esr1f/f (Ac a-tubulin, ARL13B (32), and TUBB4 (33)) were not and Wnt7aCre/+;Esr1f/f animals at 0.5 dpc showed that significantly different between Esr1f/f and Wnt7aCre/+; 34% (21/61) of WNT/CTNNB1 signaling-related genes Esr1f/f oviducts (Fig. 2G, H). In addition to overall ciliary were differentially expressed in the absence of epithelial morphology and protein levels, we assessed the length of the ESR1 (Table 1). In addition, we validated altered cilia stalk, because a previous finding indicated that normal expression of transcripts that encompassed various

ESTROGEN RECEPTOR-a IN TRANSPORT OF EMBRYOS 1599 Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 Figure 2. Loss of ESR1 in the oviductal epithelial cells leads to impaired embryo transport. A) IHC analysis of ESR1 expression in the infundibulum, ampulla, and isthmus in control (Esr1f/f) and Wnt7aCre/+;Esr1f/f oviducts. B) Percentage of wild-type embryos retrieved from the oviduct and the uterus in Esr1f/f and Wnt7aCre/+;Esr1f/f females at 3.5 dpc (n=5–6 recipients/genotype; total viable embryos retrieved = 18 from Esr1f/f and 15 from Wnt7aCre/+;Esr1f/f recipients). Ovd; oviduct. C, D) Expression of proliferative (Ki67) marker using IHC analysis (C)andquantification of Ki67-positive epithelial cells (D)inEsr1f/f and Wnt7aCre/+;Esr1f/f oviducts at 0.5 dpc (n=4–5 mice/genotype, graphs represent means 6 SEM). Data represent 14,102 epithelial cells counted in Esr1f/f and 11,991 cells in Wnt7aCre/+;Esr1f/f. E, F) H&E staining (E) and expression of ciliary (Ac a-tubulin) marker by using IHC analysis (F)inEsr1f/f and Wnt7aCre/+;Esr1f/f oviducts at 0.5 dpc. G, H) Immunoblot analysis (G) and quantification (H) of whole oviducts, indicating the levels of ciliary proteins including Ac a-tubulin, ARL13B, and TUBB4 normalized to b-actin in Esr1f/f and Wnt7aCre/+;Esr1f/f animals at 0.5 dpc (n=3 mice/genotype). I) Ciliary length measured from the infundibulum and ampulla regions of the oviduct sections stained with Ac a-tubulin in Esr1f/f and Wnt7aCre/+;Esr1f/f oviducts at 0.5 dpc. Data represent means f/f Cre/+ f/f f/f 6 SEM from a total of 3100 cells in Esr1 and 2019 cells in Wnt7a ;Esr1 oviducts (n=4 mice/genotype). *P , 0.05 vs. Esr1 , by unpaired Student’s t test. J ) CBF was measured from each ciliated cell in ex vivo cultures of Esr1f/f and Wnt7aCre/+;Esr1f/f oviducts collected at 0.5 dpc and presented as Hz in a box-and-whisker plot. Number of cells at each designated frequency was also represented in a histogram. Data from 168–469 epithelial cells were recorded from each animal. *P , 0.05 vs. Esr1f/f (n=3 mice/genotype), Mann-Whitney U test. K) Levels of Kif3a, Kif6, Kif9, Kif19,andFoxj1 transcripts in the Esr1f/f and Wnt7aCre/+;Esr1f/f oviducts at 0.5 dpc (n=4–5mice/genotype).*P , 0.05 vs Esr1f/f, unpaired Student’s t test.

WNT/CTNNB1 components using qRT-PCR analysis, contrast, however, in the Wnt7aCre/+;Esr1f/f oviducts, including WNT ligands (Wnt7a and Wnt7b) and a WNT CTNNB1 was also localized at the apical membrane of receptor (Frizzled 10 or Fzd10), a transcription factor the epithelial cells. This result suggests that the tran- (Tcf3), and CTNNB1 direct target genes (Cdh1 and script is not affected, but the intracellular localization of Fgh18). All transcripts were significantly altered, except CTNNB1 protein is disrupted by the lack of ESR1 sig- Tcf3 and Ctnnb1, in the absence of epithelial ESR1 ex- naling in the oviductal epithelial cells. pression (Fig. 3A). Although the overall transcript level of To further assess the impact of the natural physiologic Ctnnb1 was not different between Esr1f/f and Wnt7aCre/+; response to ovarian hormones on WNT/CTNNB1 sig- Esr1f/f oviducts, the protein expression pattern of CTNNB1 naling in the various regions of the oviduct, we collected differed (Fig. 3B). IHC analysis showed that the localiza- and analyzed RNA from the 3 different regions of oviducts tion of CTNNB1 was restricted only to the lateral (arrow- from wild-type animals at each stage of the estrous cycle heads), not apical (arrows), membranes of epithelial cells in [diestrus (lowest E2 level), proestrus (highest E2 level), f/f both the ampulla and the isthmus in Esr1 animals. In metestrus, and estrus] by RT-PCR. It is known that E2

1600 Vol. 31 April 2017 The FASEB Journal x www.fasebj.org LI ET AL. Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 TABLE 1. ESR1 regulates expression of canonical and noncanonical CTNNB1 is exclusively expressed in the WNT signaling-related genes epithelial cells of human fallopian tubes

Gene Fold change P To determine the appropriateness of our experimental mouse model to human studies and whether CTNNB1 Extracellular factor Dkk1 22.27 0.0031 signaling may have a clinical impact on cilia function in Dkk3 22.81 0.0006 women as well as mice, we next assessed the localization of Fzd3 2.11 0.0002 CTNNB1 protein in ciliated epithelial cells from human Fzd10 4.71 0.0001 fallopian tubes. We found that CTNNB1 was especially Wnt5b 1.54 0.0057 enriched in the lateral membrane of ciliated epithelial cells, Wnt7a 7.05 0.0028 as in the mouse samples (Fig. 4), and was not detected in 2 Wnt7b 2.01 0.0058 stromal and muscle cells. Epithelial cells also had positive Wnt11 1.76 0.0030 a Wnt16 21.73 0.0006 staining for the cilia marker Ac -tubulin. These observa- CTNNB1 regulator tions indicate that CTNNB1 may function similarly in Ppp2r2c 21.88 0.0003 mouse oviduct and human fallopian tube cilia, thereby Transcription factor having the potential of impacting human health and fer- Sox9 22.65 0.0083 tility if dysfunctional. Tcf1 2.43 0.0032 Tcf3 1.63 0.0026 CTNNB1 target gene 2 Inhibition of the WNT/CTNNB1 signaling Cd44 2.66 0.0027 pathway disrupts embryo transport ex vivo Cdh1 21.55 0.0054 Ephb2 1.82 0.0061 Fgf18 2.99 0.0003 Because we observed that lacking ESR1 in the oviductal Id2 1.86 0.0014 epithelial cells caused an aberrant expression of transcripts Lgr5 3.40 0.0005 that impact WNT/CTNNB1 signals, as well as altered Rhou 1.53 0.0017 2 localization of CTNNB1 protein, we hypothesized that Ppard 7.26 8.4E 07 ESR1-mediated E2 responses during embryo transport are Cre/+ f/f f/f mediated by WNT/CTNNB1 signaling. To directly eval- Fold change, Wnt7a ;Esr1 vs. Esr1 . uate whether the CTNNB1 signaling pathway is required for embryo transport, we employed an ex vivo oviduct culture model to compare transport distances of embryos increases the expression of WNTs and CTNNB1 but de- in the presence or absence of the WNT inhibitor IWP2 (40). creases secreted frizzled-related protein (SFRP)-2 (a neg- Oviducts were collected at 0.5 dpc with zygotes inside ative regulator of WNT ligands in the female reproductive (Fig. 5A). Oviducts were uncoiled, placed onto the tract) in mouse uteri (37), and this E -induced expression is 2 Transwell insert in a 24-well plate, and incubated with necessary for uterine cell growth. However, little is known DMSO (vehicle control) or 5 mMIWP2for3consecutive regarding E regulation of CTNNB1 in oviducts, except for 2 days. Subsequently, the distances that the embryos trav- a single study of an avian model (38). Ctnnb1 transcript eled were measured as a percentage of the distance be- was highly expressed in the infundibulum (where ciliated tween the AIJ and the UTJ. In a DMSO-treated control, the cells are most abundant), gradually decreased in the am- embryos were able to travel ;23.5 6 3.4% of the oviductal pulla, and was detected at minimal levels in the isthmus length, whereas embryos in oviducts treated with IWP2 (Fig. 3C). However, Ctnnb1 expression did not appear to traveled a significantly shorter distance (7.5 6 4.3% of the differ between the different stages of the estrous cycle, oviductal length) (Fig. 5B). To ensure that the decreased indicating ovarian hormones do not preferentially regu- transport observed in the IWP2 treatment group is related late oviductal Ctnnb1 in a stage specific manner. The to the specific effect of the treatment, we confirmed that the negative regulator of CTNNB1 Chibby1 (Cby1)(39)also level of the CTNNB1 target gene Axin2 was significantly showed an expression pattern similar to that of Ctnnb1. lower in the presence of IWP2 compared to the DMSO- In contrast, the expression of Sfrp2 was stage specific treated control (Fig. 5C). In addition, the embryos retrieved and highly expressed in the ampulla and was signifi- after 3 d of ex vivo oviduct culture were at similar de- cantly increased during proestrus and estrus when velopmental stages in both treatment groups, suggesting compared to diestrus. Expression of Axin2 and Cdh1,direct that the treatment did not result in toxicity (Fig. 5D). These CTNNB1 targets, were significantly increased in the am- findings show that inhibition of WNT/CTNNB1 signaling pulla region at proestrus compared to diestrus. Similarly, the reduces the oviduct’s ability to transport embryos. expression of Fzd9 and Wnt2 was significantly elevated in the ampulla at proestrus.Wnt7a was highly expressed in the isthmus region compared to the infundibulum and ampulla, Selective ablation of CTNNB1 in ciliated and its expression did not change at any stage of the estrous epithelial cells does not disrupt the embryo cycle. Wnt7b levels were highest in the infundibulum and transport in vivo gradually decreased in the ampulla and isthmus. These re- sults indicate that the expression of some of the WNT/ Several transcripts in the canonical WNT/CTNNB1 CTNNB1 signaling transcripts are regulated in a physio- pathway are altered in the absence of epithelial ESR1 logic-region–specific and estrous-cycle–selective manner. (Table 1). Therefore, to determine whether the canonical

ESTROGEN RECEPTOR-a IN TRANSPORT OF EMBRYOS 1601 Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 Figure 3. Expression of transcripts (A) and protein (B) involved in the WNT/CTNNB1 signaling pathway in Esr1f/f and Wnt7aCre/+; Esr1f/f oviduct at 0.5 dpc and at each stage of the estrous cycle (C) in the infundibulum, ampulla, and isthmus of wild-type oviducts. Representative images (B) show CTNNB1 IHC staining in Esr1f/f and Wnt7aCre/+;Esr1f/f oviducts at 0.5 dpc. Arrows: difference between the expression of CTNNB1 at the apical membrane of epithelial cells in Esr1f/f and Wnt7aCre/+;Esr1f/f oviducts; arrowheads: expression of CTNNB1 at the lateral membrane of the epithelial cells. Expression values determined by qRT-PCR (Wnt7a, Wnt7b, Fzd10, Tcf3, Cdh1, Fgf18, Ctnnb1, Cby1, Sfrp2, Axin2, Cdh1, Fz9, and Wnt2) were normalized to Rpl7 # f/f (n=4–5 mice/genotype). Graphs represent means 6 SEM. P , 0.05 vs. Esr1 , unpaired Student’s t test; *P , 0.05, **P , 0.01 vs. the expression value within the same region of the oviduct at diestrus (lowest circulating E2).

WNT/CTNNB1 pathway within the ciliated cells is addition, the total numbers of embryos per dam as well as downstream of E2 action and mediates embryo transport the development of embryos were comparable between through the oviduct in vivo, we generated a conditional Foxj1Cre+;Ctnnb1f/f and Ctnnb1f/f females (Fig. 6C, D). knockout mouse model where Ctnnb1 was selectively Further investigation revealed that, overall, ciliated epi- ablated in ciliated epithelial cells by crossing Foxj1Cre+ to thelial cells that were positive for Ac a-tubulin staining did Ctnnb1f/f animals. As CTNNB1 is the central protein for not appear to be different in Foxj1Cre+;Ctnnb1f/f compared canonical WNT/CTNNB1 signal, ablation of CTNNB1 to Ctnnb1f/f oviducts (Fig. 6E). Collectively, our findings would disrupt the whole signaling cascade. First, we suggest that lacking CTNNB1 in the ciliated epithelial cells confirmed that CTNNB1 protein is deleted specifically in of the oviduct does not disrupt the transport or the de- ciliated epithelial cells of the oviduct. In Ctnnb1f/f controls, velopment of the embryos, indicating that CTNNB1 in the CTNNB1 was detected in the oviduct epithelium in- secretory oviductal epithelial cells may induce the signals cluding both secretory and ciliated cells in ampulla and allowing embryo transport, summarized as a working isthmus, as well as the epithelial cells of the uterus (Fig. model in Fig. 7. 6A). CTNNB1 staining was mainly found in the lateral plasma membrane of epithelial cells, which is similar to the pattern detected in human fallopian tubes (Fig. 4). In the DISCUSSION Foxj1Cre+;Ctnnb1f/f oviduct, CTNNB1 was not detected in ciliated epithelial cells but was retained in secretory epi- In this study, we showed that ESR1 was expressed in all thelial cells (Fig. 6A). Note that CTNNB1 deletion mainly cell layers of the mouse oviduct, whereas ESR2 was spe- took place in the ampulla of the oviduct, as the isthmus cifically found in cilia of multiciliated epithelial cells. region has few ciliated cells, as indicated by Ac a-tubulin However, a global loss of Esr2 had no impact on the em- IHC (Fig. 1A). bryo transport function of the oviduct. On the contrary, We next evaluated whether the deletion of Ctnnb1 in selective ablation of Esr1 in the oviductal epithelial cells led ciliated epithelial cells impairs the transport of embryos by to a transport defect, in part because of an overall re- retrieving them from the oviduct and the uterus at 3.5 dpc. duction of CBF, possibly caused by an increase in ciliary There was no significant difference in the number of em- length. We also determined that expression of WNT/ bryos retrieved from the uterus and oviduct between CTNNB1 signaling-related genes in the oviduct was reg- Cre+ f/f f/f Foxj1 ;Ctnnb1 and Ctnnb1 females (Fig. 6B). In ulated by E2. Recent findings indicate that the expression

1602 Vol. 31 April 2017 The FASEB Journal x www.fasebj.org LI ET AL. Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 female reproductive tracts. Recently, a mouse model 2 2 lacking kinesin superfamily (Kif19 / ) showed that cilia of 2 2 Kif19 / oviduct are doubled in length compared with the control littermates (34). As KIF19 is a - depolymerizing kinesin expressed at the ciliary tips, without KIF19, the depolymerizing process is disrupted, leading to an increase in ciliary length (34). The increased 2 2 ciliary length in Kif19 / mice contributes to defective ciliary beating, resulting in ineffective fluid flow in the oviduct. In our mouse model, we found that the absence of epithelial ESR1 in the oviduct causes decreases in Kif19, Kif9,andFoxj1 transcripts. Thus, it is possible that blunted expression of Kif19 and Kif9 could cause the in- creasedciliarylengthandaberrantCBFweobservedin Wnt7aCre/+;Esr1f/f oviducts, leading to impaired embryo transport. Although the ciliary length is increased in the in- fundibulum region of Wnt7aCre/+;Esr1f/f oviducts, the Figure 4. Representative images of IHC staining for CTNNB1 levels of ciliary proteins, including Ac a-tubulin, ARL13B, a and Ac -tubulin in the epithelial cells of human fallopian and TUBB4, tended to be augmented but not significantly tubes (n=3 individuals). different from those of Esr1f/f oviducts. This finding indi- cates that the observed 14% increase of ciliary length in of CTNNB1 is increased in the ampulla region of fallopian the infundibulum (2.61 mminEsr1f/f vs. 2.98 mmin tubes of patients with diseases, such as salpingitis, and Wnt7aCre/+;Esr1f/f oviduct) does not alter the levels of ectopic pregnancy (41). Together with our findings, these overall ciliary protein expression in the whole oviducts. observations suggest that aberrant expression of CTNNB1 in the fallopian tubes could result in ectopic pregnancy. In the mouse, inhibition of WNT/CTNNB1 signaling dis- rupted the embryo transport in an ex vivo oviduct culture. However, conditional removal of CTNNB1 in oviductal ciliated epithelial cells had no effect on the embryo trans- port through the oviduct. Thus, our findings demonstrate expression WNT/CTNNB1 in oviductal ciliated epithelial cells is not required for embryo transport but rather ex- pression of epithelial ESR1 in the oviduct is essential for the proper ciliary function of the oviduct. It has been reported that ESR2, which is detected in oviductal cilia, modulates the expression of calbindin- D28K (20), a signaling molecule critical for Ca2+-regulated ciliary beating. Our findings, however, showed that de- letion of Esr2 had no impact on the transport of embryos. This suggests that ESR2 signaling, and its downstream pathways, appear not to be involved in the regulation of ciliary beating of the oviduct during the transport of embryos. Previous studies have demonstrated that pharmaco- logical inhibition of ESR1 function using an ESR antagonist suppresses the proliferation of oviductal ciliated cells in neonatal mice (17). In addition, extensive loss of ciliated Figure 5. Inhibition of WNT/CTNNB1 signaling pathway leads cells and shortening of ciliary length are observed in rab- to an attenuation of the ex vivo embryo transport. A) bits after ovariectomy (42), suggesting that loss of ovarian Procedure for ex vivo oviduct culture in vehicle (DMSO) or IWP2 at a dose of 5 mM, for 3 consecutive days. B) The distance hormones dramatically affects ciliated cell morphology the embryos traveled from AIJ to UTJ in the presence or and growth. Moreover, E2 induces expression of Foxj1 absence of IWP2. The distances were measured from AIJ to (a central regulator of the motile ciliogenesis) in the mouse UTJ and calculated as a percentage of the distance, in which oviduct and rhesus macaque fallopian tubes (17, 43). Ad- the embryos that travel to the UTJ are 100%. C) Axin2 (target ditionally, the number of ciliated epithelial cells in the ef- gene of the canonical WNT/CTNNB1 pathway) transcript in the oviducts that were cultured in the presence or absence of ferent duct in male is also regulated by E2 signaling as a 2/2 IWP2 for 3 consecutive days (n=4–5 mice/group, graphs global deletion of Esr1 (Esr1 )causesadecreaseinthe , +/+ represent means SEM).*P 0.05 vs. vehicle control, unpaired number of motile cilia compared to Esr1 mice (44). Student’s t test. D) Representative images of embryos flushed These findings indicate that E2 regulates the turnover of from the oviducts that were treated with DMSO or IWP2 for 3 ciliated epithelial cells through ESR1 in both male and consecutive days.

ESTROGEN RECEPTOR-a IN TRANSPORT OF EMBRYOS 1603 Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 Figure 6. Loss of CTNNB1 in ciliated epithelial cells does not affect the transport of embryos within the oviduct. A) IHC analysis of CTNNB1 in the ampulla, isthmus, and uterus of control (Ctnnb1f/f) and Foxj1Cre+;Ctnnb1f/f females. Note that the expression of CTNNB1 was not detected in the ciliated epithelial cells of Foxj1Cre+;Ctnnb1f/f females. B) Percentage of the embryos (both morulae and blastocysts) retrieved from the oviduct and the uterus in Ctnnb1f/f and Foxj1Cre+;Ctnnb1f/f females at 3.5 dpc. C) Total f/f Cre+ f/f number of embryos in Ctnnb1 and Foxj1 ;Ctnnb1 females at 3.5 dpc. Graphs represent means 6 SEM (n=6–9 mice/ genotype). D) Representative images of embryos flushed from the oviducts and uterus of Ctnnb1f/f and Foxj1Cre+;Ctnnb1f/f females at 3.5 dpc. E) Ciliated epithelial cells of Ctnnb1f/f and Foxj1Cre+;Ctnnb1f/f oviducts indicated by Ac a-tubulin IHC.

Several studies reported that exogenous treatment with CTNNB1/LEF1 complex in rat uterine and MCF-7 cells E2 can increase or decrease the embryo transport rate (56, 57). Inversely, uterine expression of the negative reg- depending on the dose administered (45–47). In rodents, ulator of WNT/CTNNB1 signaling SFRP2 is suppressed the effects of E2 on embryo transport in the oviduct have by E2 treatment (37). In this present study, we found that been reported to be mediated through both genomic and expression of many WNT/CTNNB1 signaling molecules nongenomic actions (48–50). Collectively, this indicates were significantly altered after deletion of ESR1 in ovi- that E2 modulates the CBF in the oviduct. In this study, we ductal epithelial cells. In addition, CTNNB1 protein was found that in the absence of ESR1 in the epithelial cells, the detected on the apical membrane of epithelial cells in the median of CBFs was significantly lower when compared absence of epithelial ESR1, whereas the apical localization to the controls. In addition, there were a number of ciliated of CTNNB1 was minimally detected in the control an- cells that beat at higher frequencies (.20 Hz), which could imals (either Esr1f/f or Ctnnb1f/f). potentially alter the net transport rate of the embryos. Hy- WNT/CTNNB1 signaling has been linked to cilia drodynamic force (or the fluid flow) is one of the factors function (58). Mice with loss of Cby, a repressor of canon- affecting the CBFs (51, 52). Several studies in rabbits and ical WNT/CTNNB1-dependent gene transcription, de- humans demonstrated that E2 increases the fluid flow in the velop sinusitis and otitis because of a marked reduction in oviduct (53, 54). Moreover, our recent findings indicate that the number of motile cilia, leading to defective mucociliary 2 2 a lack of ESR1 in the epithelial cells leads to an alteration in clearance (39, 59). Cby / males and females are subfertile secretory proteins produced by the oviduct, including (39). These results suggest that dysregulation of the WNT/ proteases and protease inhibitors, resulting in an embryonic CTNNB1 pathway can result in defective motile cilia. Be- 2 2 death within the oviduct (22). Taken together, these data cause global loss of CTNNB1 (Ctnnb1 / )isembryonic imply that the absence of ESR1 in the oviductal epithelial lethal due to a gastrulation defect (60), this mouse model cells could potentially disrupt fluid secretion, which would cannot be used to study CTNNB1 function in the oviduct. subsequently alter the transport of the embryos as a result of Similarly, mice carrying a tissue-specific deletion of cumulative loss of fluid flow and irregular CBFs. CTNNB1 in the stromal and muscle cell layers of female Upon binding of a WNT ligand to its receptor, FZD, reproductive tract (using Amhr2Cre/+;Ctnnb1f/f animals) CTNNB1 translocates into the nucleus and stimulates cannot be used for such studies because of oviduct de- target gene expression through interactions with the T-cell velopmental defects, including partial oviduct formation, factor/lymphocyte enhancer factor-1 transcription factor failure of coiling, and lack of mucosal folding of the epi- (55). E2 increases expression of WNT4, WNT5A, and thelial layer (61, 62). We generated a conditional-knockout CTNNB1 in mouse uterine epithelial cells (38) and induces mouse line lacking Ctnnb1 selectively in ciliated epithelial the nuclear translocation of CTNNB1 and formation of the cells to test whether CTNNB1 is functionally required for

1604 Vol. 31 April 2017 The FASEB Journal x www.fasebj.org LI ET AL. Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 Figure 7. Working model illustrates E2 and CTNNB1 signals in the oviductal ciliated (blue) and secretory (orange) epithelial cells regulating embryo transport. A) In the presence of ESR1, WNT/CTNNB1 is normally expressed, leads to a normal embryo transport, regardless of ESR2 expression in the ciliated epithelial cells. B) In the absence of ESR1, expression of WNT/CTNNB1, Kif9, Foxj1 are aberrantly expressed, ciliary length and beating are altered, causing impaired embryo transport. C) In the presence of ESR1, ex vivo attenuation of WNT/CTNNB1 signals in oviductal epithelial cells disrupts embryo transport, indicating that the WNT/CTNNB1 pathway is a potential downstream mediator of E2-regulation of embryo transport through the oviduct. fi D) Loss of CTNNB1 speci cally in the ciliated epithelial cells does not affect oviductal embryo transport, indicating that E2 modulates WNT/CTNNB1 signals in secretory epithelial cells, and also acts through the noncanonical WNT signaling pathway or other downstream pathways in the ciliated epithelial cells, to control the embryo transport. the function of ciliated cells during embryo transport. To use of ESR antagonists or via disruptionofESR1itselfinthe our surprise, there was no embryo transport defect in the epithelial cells of the oviduct impairs embryo transport, absence of Ctnnb1 in ciliated cells, which suggested that increasing the risk of ectopic pregnancy. CTNNB1 intrinsic to ciliated epithelial cells is not required In summary, our study provides evidence that E2 during embryo transport. However, CTNNB1 is still pre- signal via ESR1 is clearly required for the embryo sent in nonciliated cells and may affect ciliated cells via transport function of the oviduct. The details of E2- paracrine signaling. This interpretation is supported by regulated molecular mechanisms in the multiciliated our ex vivo culture results clearly showing that inhibition of epithelial cells and secretory cells of the oviduct will WNT/CTNNB1 signaling disrupted the embryo trans- require further analysis. port.Itislikelythatthetransportdefectobservedintheex vivo model resulted from the inhibitory effects on both ACKNOWLEDGMENTS secretory and ciliated epithelial cell functions. It was pre- viously reported that primary cilia are present in the The authors thank Sylvia Hewitt and Karina Rodriguez [U.S. nonciliated epithelial cells of human fallopian tubes (63) National Institutes of Health, National Institute of Environmental and WNT/CTNNB1 signaling is crucial for primary cilia Health Sciences (NIEHS)] for critical reading of this manuscript, function (64). Therefore, the function of nonciliated (se- Carmen Williams (NIEHS) for constructive advice for the cretory) epithelial cells could also be affected as IWP2 also project, Page Myers (NIEHS) for the oviduct transfer procedure, inhibits noncanonical WNT signaling. Thus, the possibility Jeff Tucker (NIEHS) for technical assistance, NIEHS Histology Core for ESR2 IHC staining, and Research Histology Core that the noncanonical WNT pathway may be involved in Facility (Department of Pathology, Stony Brook University the regulation of cilia and nonmotile function in the ex vivo School of Medicine) for the human fallopian tube samples. This oviduct culture has to be considered. work is supported by the start-up fund from College of Veterinary A growing number of case reports suggest that the Medicine (Washington State University; to W.W.) and Intramural prolonged use of clomiphene citrate (Clomid; Sanofi research Grant 1ZIAES70065 from NIEHS (to K.S.K.). The fl Aventis, Gentilly, France) is linked with bilateral ectopic authors declare no con icts of interest. pregnancy (65–67). Clomid has been widely used more than half a century by fertility clinics to induce ovulation in AUTHOR CONTRIBUTIONS infertile women (65). It is a selective estrogen receptor modulator (SERM), and acts by binding to the ESRs and W. Winuthayanon designed the research; W. Winuthayanon, ’ blocking E2 action (68), increasing the release of follicle S. Li, and S. R. S. O Neill performed research and analyzed stimulating hormone and luteinizing hormone (69). The data; S. Li and W. Winuthayanon wrote the paper; and K.-I. cause of bilateral ectopic pregnancy from the use of Clo- Takemaru,Y.Zhang,M.J.Holtzman,andK.S.Korach mid has not been investigated, but we can speculate contributed new protocols, samples, and mouse lines crucial whether it is the result of disrupted E2/ESR1 signaling in for the study. the fallopian tubes. Findings in rats suggest that animals treated with Clomid showed delayed oocyte transport within the oviduct (70). In this study, we showed that REFERENCES disruption of E signal in the epithelial cells of the mouse 2 1. Dirksen, E. R., and Stair, P. (1972) Ciliary activity in the mouse oviduct oviduct causes an embryo transport defect. Together, as studied by transmission and scanning electron microscopy. Tissue – these findings indicate that aberrant E2 signaling, either via Cell 4,389 403

ESTROGEN RECEPTOR-a IN TRANSPORT OF EMBRYOS 1605 Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 2. Gaddum-Rose, P., and Blandau, R. J. (1976) Comparative 26. Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., observations on ciliary currents in mammalian oviducts. Biol. Pietzsch, T., Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B., Reprod. 14, 605–609 Tinevez, J. Y., White, D. J., Hartenstein, V., Eliceiri, K., Tomancak, P., 3. Talbot, P., Geiske, C., and Knoll, M. (1999) Oocyte pickup by the and Cardona, A. (2012) Fiji: an open-source platform for biological- mammalian oviduct. Mol. Biol. Cell 10,5–8 image analysis. Nat. Methods 9,676–682 4. Blandau, R. J., and Gaddum-Rosse, P. (1974) Mechanism of sperm 27. Westwood, F. R. (2008) The female rat reproductive cycle: a practical transport in pig oviducts. Fertil. Steril. 25,61–67 histological guide to staging. Toxicol. Pathol. 36, 375–384 5. Halbert, S. A., Becker, D. R., and Szal, S. E. (1989) Ovum transport in 28.Rizos,D.,Ramirez,M.A.,Pintado,B.,Lonergan,P.,and the rat oviductal ampulla in the absence of muscle contractility. Biol. Gutierrez-Adan, A. (2010) Culture of bovine embryos in Reprod. 40,1131–1136 intermediate host oviducts with emphasis on the isolated mouse 6. Halbert, S. A., Tam, P. Y., Adams, R. J., and Blandau, R. J. (1976) An oviduct. Theriogenology 73, 777–785 analysis of the mechanisms of egg transport in the ampulla of the 29. Humphrey, K. W. (1968) Observations on transport of ova in the rabbit oviduct. Gynecol. Invest. 7,306–320 oviductofthemouse.J. Endocrinol. 40, 267–273 7. Halbert, S. A., Tam, P. Y., and Blandau, R. J. (1976) Egg transport in 30. Krege, J. H., Hodgin, J. B., Couse, J. F., Enmark, E., Warner, M., the rabbit oviduct: the roles of cilia and muscle. Science 191, 1052–1053 Mahler,J.F.,Sar,M.,Korach,K.S.,Gustafsson,J.A.,andSmithies,O. 8. Eliasson, R., Mossberg, B., Camner, P., and Afzelius, B. A. (1977) The (1998) Generation and reproductive phenotypes of mice lacking immotile-cilia syndrome: a congenital ciliary abnormality as an etio- estrogen receptor beta. Proc. Natl. Acad. Sci. USA 95, 15677–15682 logic factor in chronic airway infections and male sterility. N. Engl. J. 31. Roblero, L. S., and Garavagno, A. C. (1979) Effect of oestradiol-17 Med. 297,1–6 beta and progesterone on oviductal transport and early development 9. Lurie,M.,Tur-Kaspa,I.,Weill,S.,Katz,I.,Rabinovici,J.,and of mouse embryos. J. Reprod. Fertil. 57,91–95 Goldenberg, S. (1989) Ciliary ultrastructure of respiratory and 32. Delling, M., DeCaen, P. G., Doerner, J. F., Febvay, S., and Clapham, fallopian tube epithelium in a sterile woman with Kartagener’s D. E. (2013) Primary cilia are specialized calcium signalling syndrome: a quantitative estimation. Chest 95, 578–581 organelles. Nature 504,311–314 10. Afzelius, B. A. (2004) Cilia-related diseases. J. Pathol. 204,470–477 33. Tichelaar, J. W., Wert, S. E., Costa, R. H., Kimura, S., and Whitsett, J. A. 11. Halbert, S. A., Patton, D. L., Zarutskie, P. W., and Soules, M. R. (1997) (1999) HNF-3/forkhead homologue-4 (HFH-4) is expressed in cili- Function and structure of cilia in the fallopian tube of an infertile ated epithelial cells in the developing mouse lung. J. Histochem. Cyto- woman with Kartagener’ssyndrome.Hum. Reprod. 12,55–58 chem. 47,823–831 12. McComb, P., Langley, L., Villalon, M., and Verdugo, P. (1986) The 34. Niwa, S., Nakajima, K., Miki, H., Minato, Y., Wang, D., and Hirokawa, oviductal cilia and Kartagener’ssyndrome.Fertil. Steril. 46,412–416 N. (2012) KIF19A is a microtubule-depolymerizing kinesin for ciliary 13. Afzelius, B. A., Camner, P., and Mossberg, B. (1978) On the function length control. Dev. Cell 23,1167–1175 of cilia in the female reproductive tract. Fertil. Steril. 29,72–74 35. Brechbuhl, H. M., Ghosh, M., Smith, M. K., Smith, R. W., Li, B., Hicks, 14. Lyons, R. A., Saridogan, E., and Djahanbakhch, O. (2006) The D. A., Cole, B. B., Reynolds,P. R., andReynolds,S.D.(2011) b-catenin reproductive significance of human Fallopian tube cilia. Hum. Reprod. dosage is a critical determinant of tracheal basal cell fate Update 12,363–372 determination. Am.J.Pathol.179,367–379 15. Allen, E. (1922) The oestrous cycle in the mouse. Dev. Dyn. 30, 36. MacDonald, B. T., Tamai, K., and He, X. (2009) Wnt/beta-catenin 297–371 signaling: components, mechanisms, and diseases. Dev. Cell 17,9–26 16. Crow, J., Amso, N. N., Lewin, J., and Shaw, R. W. (1994) Morphology 37. Hou, X., Tan, Y., Li, M., Dey, S. K., and Das, S. K. (2004) Canonical and ultrastructure of fallopian tube epithelium at different stages of Wnt signaling is critical to estrogen-mediated uterine growth. Mol. the menstrual cycle and menopause. Hum. Reprod. 9,2224–2233 Endocrinol. 18,3035–3049 17. Okada, A., Ohta, Y., Brody, S. L., Watanabe, H., Krust, A., Chambon, 38. Bae,S.M.,Lim,W.,Jeong,W.,Lee,J.Y.,Kim,J.,Han,J.Y.,Bazer,F.W., P., and Iguchi, T. (2004) Role of foxj1 and estrogen receptor alpha in and Song, G. (2014) Hormonal regulation of beta-catenin during ciliated epithelial cell differentiation of the neonatal oviduct. J. Mol. development of the avian oviduct and its expression in epithelial cell- Endocrinol. 32,615–625 derived ovarian carcinogenesis. Mol. Cell. Endocrinol. 382,46–54 18. Teilmann, S. C., Clement, C. A., Thorup, J., Byskov, A. G., and 39. Voronina,V.A.,Takemaru,K.,Treuting,P.,Love,D.,Grubb,B.R., Christensen, S. T. (2006) Expression and localization of the Hajjar, A. M., Adams, A., Li, F. Q., and Moon, R. T. (2009) Inactivation progesterone receptor in mouse and human reproductive organs. of Chibby affects function of motile airway cilia. J. Cell Biol. 185, J. Endocrinol. 191,525–535 225–233 19. Okada, A., Ohta, Y., Inoue, S., Hiroi, H., Muramatsu, M., and Iguchi, 40. Chen,B.,Dodge,M.E.,Tang,W.,Lu,J.,Ma,Z.,Fan,C.W.,Wei,S., T. (2003) Expression of estrogen, progesterone and androgen Hao, W., Kilgore, J., Williams,N. S., Roth, M. G., Amatruda, J. F., Chen, receptors in the oviduct of developing, cycling and pre- C., and Lum, L. (2009) Small molecule-mediated disruption of Wnt- implantation rats. J. Mol. Endocrinol. 30, 301–315 dependent signaling in tissue regeneration and cancer. Nat. Chem. 20. Shao, R., Weijdegard,˚ B., Fernandez-Rodriguez, J., Egecioglu, E., Zhu, Biol. 5,100–107 C., Andersson, N., Thurin-Kjellberg, A., Bergh, C., and Billig, H. 41. Li,P.,Zhu,W.J.,Ma,Z.L.,Wang,G.,Peng,H.,Chen,Y.,Lee,K.K., (2007) Ciliated epithelial-specific and regional-specific expression and Yang, X. (2013) Enhanced beta-catenin expression and in- and regulationof the estrogen receptor-beta2 in the fallopian tubes of flammation are associated with human ectopic tubal pregnancy. immature rats: a possible mechanism for estrogen-mediated transport Hum. Reprod. 28,2363–2371 process in vivo. Am. J. Physiol. Endocrinol. Metab. 293,E147–E158 42. Rumery, R. E., and Eddy, E. M. (1974) Scanning electron microscopy 21. Winuthayanon, W., Hewitt, S. C., Orvis, G. D., Behringer, R. R., and of the fimbriae and ampullae of rabbit oviducts. Anat. Rec. 178,83–101 Korach, K. S. (2010) Uterine epithelial estrogen receptor a is 43. Ohm, L. A., Shirendeb, U., Keator, C. S., Mah, K., Rothlein, S. R., and dispensable for proliferation but essential for complete biological and Slayden, O. D. (2009) Estrogen stimulates expression of FOXJ1 in the biochemical responses. Proc. Natl. Acad. Sci. USA 107, 19272–19277 rhesus macaque fallopian tube. Fertil. Steril. 92, S116 22. Winuthayanon, W., Bernhardt, M. L., Padilla-Banks, E., Myers, P. H., 44. Hess, R. A., Bunick, D., Lubahn, D. B., Zhou, Q., and Bouma, J. (2000) Edin, M. L., Lih, F. B., Hewitt, S. C., Korach, K. S., and Williams, C. J. Morphologic changes in efferent ductules and epididymis in estrogen (2015) Oviductal estrogen receptor a signaling prevents protease- receptor-alpha knockout mice. J. Androl. 21,107–121 mediated embryo death. eLife 4, e10453 45. Roblero,L.S., Fernandez,´ O., and Croxatto, H. B. (1987)The effect of 23. Deroo, B. J., Rodriguez, K. F., Couse, J. F., Hamilton, K. J., Collins, J. B., RU486 on transport, development and implantation of mouse Grissom,S.F.,andKorach,K.S.(2009)Estrogenreceptorbetais embryos. Contraception 36,549–555 required for optimal cAMP production in mouse granulosa cells. Mol. 46. Yang, Y. Q., and Wu, J. T. (1990) RU 486 interferes with egg transport Endocrinol. 23,955–965 and retards the in vivo and in vitro development of mouse embryos. 24. Zhang, Y., Huang, G., Shornick, L. P., Roswit, W. T., Shipley, J. M., Contraception 41,551–556 Brody, S. L., and Holtzman, M. J. (2007) A transgenic FOXJ1-Cre 47. Greenwald, G. S. (1967) Species differences in egg transport in system for gene inactivation in ciliated epithelial cells. Am.J.Respir.Cell response to exogenous estrogen. Anat. Rec. 157,163–172 Mol. Biol. 36,5159–51 48.Nishimura,A.,Sakuma,K.,Shimamoto,C.,Ito,S.,Nakano,T., 25. Winuthayanon, W., Hewitt, S. C., and Korach, K. S. (2014) Uterine Daikoku,E.,Ohmichi,M.,Ushiroyama,T.,Ueki,M.,Kuwabara,H., epithelial cell estrogen receptor alpha-dependent and -independent Mori, H., and Nakahari, T. (2010) Ciliary beat frequency controlled by genomic profiles that underlie estrogen responses in mice. Biol. oestradiol and progesterone during ovarian cycle in guinea-pig fal- Reprod. 91,110 lopian tube. Exp. Physiol. 95 ,819–828

1606 Vol. 31 April 2017 The FASEB Journal x www.fasebj.org LI ET AL. Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 49. Orihuela, P. A., and Croxatto, H. B. (2001) Acceleration of oviductal mechanics in mice deficient in the Wnt/b-catenin antagonist transport of oocytes induced by estradiol in cycling rats is mediated by Chibby. PLoS One 5,e13600 nongenomic stimulation of protein phosphorylation in the oviduct. 60. Haegel, H., Larue, L., Ohsugi, M., Fedorov, L., Herrenknecht, K., and Biol. Reprod. 65,1238–1245 Kemler, R. (1995) Lack of beta-catenin affects mouse development at 50. Orihuela, P. A., Parada-Bustamante, A., Cort´es, P. P., Gatica, C., and gastrulation. Development 121 , 3529–3537 Croxatto, H. B. (2003) Estrogen receptor, cyclic adenosine 61. Hernandez Gifford, J. A., Hunzicker-Dunn, M. E., and Nilson, J. H. monophosphate, and protein kinase A are involved in the (2009) Conditional deletion of beta-catenin mediated by Amhr2cre nongenomic pathway by which estradiol accelerates oviductal in mice causes female infertility. Biol. Reprod. 80, 1282–1292 oocyte transport in cyclic rats. Biol. Reprod. 68,1225–1231 62. Deutscher, E., and Hung-Chang Yao, H. (2007) Essential roles of 51. Guirao, B., Meunier, A., Mortaud, S., Aguilar, A., Corsi, J. M., Strehl, mesenchyme-derived beta-catenin in mouse M¨ullerian duct mor- L.,Hirota,Y.,Desoeuvre,A.,Boutin,C.,Han,Y.G.,Mirzadeh,Z., phogenesis. Dev. Biol. 307,227–236 Cremer, H., Montcouquiol, M., Sawamoto, K., and Spassky, N. (2010) 63. Hagiwara, H., Ohwada, N., Aoki, T., Suzuki, T., and Takata, K. (2008) Coupling between hydrodynamic forces and planar cell polarity The primarycilia of secretorycells in the human oviduct mucosa. Med. orients mammalian motile cilia. Nat. Cell Biol. 12, 341–350 Mol. Morphol. 41,193–198 52. Yang, X., Dillon, R. H., and Fauci, L. J. (2008) An integrative 64. Goetz, S. C., and Anderson, K. V. (2010) The primary cilium: a computational model of multiciliary beating. Bull. Math. Biol. 70, signalling centre during vertebrate development. Nat. Rev. Genet. 11, 1192–1215 331–344 53. Leese, H. J. (1988) The formation and function of oviduct fluid. 65. Khong, S. Y., and Dimitry, E. D. (2005) Bilateral tubal ectopic J. Reprod. Fertil. 82,843–856 pregnancy after clomiphene induction. J. Obstet. Gynaecol. 25,611–612 54. Lippes, J., Krasner, J., Alfonso, L. A., Dacalos, E. D., and Lucero, R. 66. Mathew, M., Saquib, S., and Krolikowski, A. (2004) Simultaneous (1981) Human oviductal fluid proteins. Fertil. Steril. 36,623–629 bilateral tubal pregnancy after ovulation induction with clomiphene 55. Behrens, J., von Kries, J. P., K¨uhl, M., Bruhn, L., Wedlich, D., citrate. SaudiMed.J.25, 2058–2059 Grosschedl, R., and Birchmeier, W. (1996) Functional interaction of 67. Harrison, K., Molloy, D., Doody, M., and Breen, T. (1993) beta-catenin with the transcription factor LEF-1. Nature 382,638–642 Clomiphene citrate and pregnancy loss. Hum. Reprod. 8,981 56. Gupta, N., Schmitt, F., Grebhardt, S., and Mayer, D. (2011) b-Catenin 68. Plouffe, L., Jr. (2000) Selective estrogen receptor modulators is a positive regulator of estrogen receptor-afunction in breast cancer (SERMs) in clinical practice. J. Soc. Gynecol. Investig. 7(1, Suppl) cells. Cancers (Basel) 3,2990–3001 S38–S46 57. Rider,V.,Isuzugawa,K.,Twarog,M.,Jones,S.,Cameron,B.,Imakawa, 69. Kerin, J. F., Liu, J. H., Phillipou, G., and Yen, S. S. (1985) Evidence for a K., and Fang, J. (2006) Progesterone initiates Wnt-beta-catenin sig- hypothalamic site of action of clomiphene citrate in women. J. Clin. naling but estradiol is required for nuclear activation and synchro- Endocrinol. Metab. 61,265–268 nous proliferation of rat uterine stromal cells. J. Endocrinol. 191, 70. Shao, R., Nutu, M., Weijdegard,˚ B., Egecioglu, E., Fernandez-Rodriguez, 537–548 J., Karlsson-Lindahl, L., Gemzell-Danielsson, K., Bergh, C., and Billig, H. 58. Wallingford, J. B., and Mitchell, B. (2011) Strange as it may seem: the (2009) Clomiphene citrate causes aberrant tubal apoptosis and many links between Wnt signaling, planar cell polarity, and cilia. Genes estrogen receptor activation in rat fallopian tube: implications for Dev. 25,201–213 tubal ectopic pregnancy. Biol. Reprod. 80,1262–1271 59. Love, D., Li, F. Q., Burke, M. C., Cyge, B., Ohmitsu, M., Cabello, J., Larson,J.E.,Brody,S.L.,Cohen,J.C.,andTakemaru,K.(2010) Received for publication October 7, 2016. Altered lung morphogenesis, epithelial cell differentiation and Accepted for publication December 19, 2016.

ESTROGEN RECEPTOR-a IN TRANSPORT OF EMBRYOS 1607 Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017 Estrogen receptor α is required for oviductal transport of embryos

Shuai Li, Sofia R. S. O'Neill, Yong Zhang, et al.

FASEB J 2017 31: 1595-1607 originally published online January 12, 2017 Access the most recent version at doi:10.1096/fj.201601128R

References This article cites 70 articles, 30 of which can be accessed free at: http://www.fasebj.org/content/31/4/1595.full.html#ref-list-1

Subscriptions Information about subscribing to The FASEB Journal is online at http://www.faseb.org/The-FASEB-Journal/Librarian-s-Resources.aspx

Permissions Submit copyright permission requests at: http://www.fasebj.org/site/misc/copyright.xhtml

Email Alerts Receive free email alerts when new an article cites this article - sign up at http://www.fasebj.org/cgi/alerts

© FASEB

Downloaded from www.fasebj.org to IP 134.121.125.9. The FASEB Journal Vol.31, No.4 , pp:1595-1607, March, 2017