| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
Stimulates the Janus Kinase-Signal Transducer and Activator of Transcription Pathway in Bovine Endometrial Epithelial Cells1
a Departments of Dairy and Poultry Sciences and
b Microbiology and Cell Science, University of Florida, Gainesville, Florida 32611
c Center for Animal Biotechnology and Genomics, Texas A&M University System Health Science Center, Texas A&M University, College Station, Texas 77843
d Department of Animal Science, University of Wyoming, Laramie, Wyoming 82071
ABSTRACT
Trophoblastic bovine interferon-tau (bIFN-
) suppresses luteolytic pulses of endometrial prostaglandin F2
(PGF2) at the time of maternal recognition of pregnancy. This results in maintenance of the corpus luteum in cattle. The hypothesis that effects of bIFN- in the endometrium were through activation of the Janus kinase (JAK)-signal transducer and activator of transcription (STAT) pathway of signal transduction was tested. Whole cell, cytosolic, and nuclear extracts from bovine endometrial cells treated with bIFN- were analyzed by immunoprecipitation, immunoblotting, and electrophoretic mobility shift assays in a series of dose- and time-dependency experiments. Bovine IFN- stimulated tyrosine phosphorylation, homo- and heterodimer formation, nuclear translocation, and DNA binding of STAT proteins 1, 2, and 3. Moreover, bIFN- induced synthesis of interferon-regulatory factor. In conclusion, bIFN- stimulates the JAK-STAT pathway in the bovine endometrium. It is proposed that activation of the JAK-STAT pathway is involved in regulating the antiluteolytic effects of bIFN-.
conceptus, mechanisms of hormone action, pregnancy, signal transduction, uterus
INTRODUCTION
In cattle, suppression of luteolytic pulses of prostaglandin (PG)F2 from the pregnant endometrium is required for maintenance of pregnancy. Presence of a conceptus at around Day 17 of pregnancy blocks production of endometrial PGF2 and thereby maintains the corpus luteum (CL) [1]. Several studies indicate that the bovine (b) conceptus secretes interferon-tau (IFN-) during the time of CL maintenance [2–4]. The bIFN- suppresses oxytocin-induced release of PGF2 in vivo [5], and in primary cultures of endometrial epithelial cells (the major source of uterine PGF2) [6, 7]. Also, uterine infusions of bIFN- extended the CL life span in cows [5]. In order to exert antiluteolytic functions, bIFN- must bind to receptors on endometrial epithelial cells and stimulate signal transduction mechanisms to suppress PGF2 pulses. Binding of bIFN- has been demonstrated in bovine endometrial cells [8]. Moreover, Binelli et al. [9, 10] and Perry et al. [11] suggested that bIFN- activates the Janus kinase (JAK)-signal transducer and activator of transcription (STAT) signal transduction pathway [12, 13] in bovine endometrial cells. The working model is that binding of bIFN- to its receptor elicits tyrosine phosphorylation of STAT proteins in the cytosol of endometrial epithelial cells. In other systems, phosphorylated STATs form homo- and heterodimers [13, 14] that translocate to the nucleus. Binding of nuclear STAT complexes to specific sequences in the regulatory region of interferon-regulated genes activates transcription [12]. It is hypothesized that bIFN--stimulated proteins inhibit one or several steps of the cascade for synthesis of PGF2 in the endometrium.
Type I interferons direct transcription of both interferon-regulatory factor (IRF)-1 and IRF-2 [15]. Interferon RF-1 directly activates expression of IRF-2 [16]. Conversely, IRF-2 negatively regulates expression of IRF-1, characterizing a yin-yang type of regulatory paradigm [15, 17]. Interferon RF-2 is a transcription repressor that inhibits transcription of genes such as the Epstein-Barr virus nuclear antigen 1 (EBNA-1) [18] and the nitric oxide synthase-2 (NOS-2) [19]. Interferon RF-2 is induced by ovine IFN- in the endometrium of sheep [20] and may suppress transcription of genes critical to the generation of PGF2 such as the bovine oxytocin receptor [21] and the ovine estrogen receptor [22] genes that contain putative IRF response elements (IREs).
The present study examined the signal transduction pathway stimulated by bIFN- in a line of bovine endometrial epithelial cells (BEND cells) that originated from Day 14 cyclic cows [23]. Specific objectives were to 1) characterize the phenotype of BEND cells and confirm their epithelial nature; 2) test the dose responsiveness of bIFN--induced phosphorylation of STAT-1 and -2; 3) study the time dependence of bIFN--induced phosphorylation of STAT-1, -2, and -3; 4) verify whether bIFN- induces nuclear translocation of STAT proteins; 5) examine the formation of STAT-STAT complexes; 6) identify the nature of nuclear proteins binding interferon stimulus response elements (ISRE) and sis-inducible elements (SIE); and 7) examine the ability of bIFN- to induce synthesis of IRF-1 in BEND cells.
MATERIALS AND METHODS
Materials
Acrylamide, N,N'-methylenebisacrylamide, SDS, and Nonidet-P40 were from BDH Laboratory Supplies (Poole, England). Bovine IFN- (200 µg/ml in 20 mM Tris-HCl, pH 8; 1.08 x 107 units of antiviral activity/mg) was a gift from Dr. Michael Roberts from the University of Missouri. Tissue culture plastic petri dishes were purchased from Corning Glass Works (Corning, NY). Polystyrene cell culture flasks were from Sarstedt, Inc. (Newton, NC). LabTek four-well chamber slides were from Nunc (Naperville, IL). Tris, Tris-HCl, NaCl, EDTA, NaF, glycerol, glycine, methanol, gelatin, Tween-20, Tris-HCl, EDTA, Hepes, ß-mercaptoethanol, 15-ml polypropylene conical tubes, 100- x 16-mm polypropylene centrifuge tubes, borosilicate Pasteur pipettes (5.75''), ammonium acetate, KCl, MgCl2, dithiothreitol (DTT), glycerol, xylene cyanol, and Whatman filter paper were purchased from Fisher Scientific (Pittsburgh, PA). Aprotinin, leupeptin, pepstatin, Na4P2O7, EGTA, Na3VO4, PMSF and protein-A agarose beads, monoclonal anti-vimentin clone V9 antibody, monoclonal anti-pan cytokeratin antibody, Hams F-12, minimal essential medium (MEM), antibiotic-antimycotic solution, insulin, D-valine, horse serum, trypsin solution, Sephadex G-10 beads, rabbit IgG, and tRNA were from Sigma Chemical Co. (St. Louis, MO). Anti-STAT-1, anti-STAT-2, anti-STAT-3, and anti-IRF-1 antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA); anti-phospho-STAT-1 and anti-phospho-STAT-3 were from New England Biolabs (Beverly, MA); and anti-phosphotyrosine (PY-20) antibody was from Transduction Laboratories (Lexington, KY). Fluorescein-conjugated secondary antibody (goat anti-rabbit IgG) was from Zymed (San Francisco, CA). Prolong antifade mounting reagent was purchased from Molecular Probes (Eugene, OR). Nitrocellulose membranes (Hybond-ECL) were from Amersham (Buckingamshire, England). Enhanced chemiluminescence (ECL) kit (Renaissance Western Blot Chemiluminescence Reagent Plus) and x-ray films were from NEN Research Products (Reflection; Boston, MA). Nonfat dried milk was from Mid-America Farms (Springfield, MO). Fetal bovine serum was acquired from Atlanta Biologicals (Norcross, GA). Bio-Rad protein assay was from Bio-Rad Laboratories (Hercules, CA). Complementary oligodeoxyribonucleotides corresponding to the ISRE (double stranded, 100 ng/µl solution) and to the SIE (sense and anti-sense strands, 100 ng/µl solutions) were obtained as a gift from Dr. Douglas Leaman (The Cleveland Clinic Foundation, OH). Sequences of the sense (5' to 3') strands were as follows: ISRE, 5'-TTTACAAACAGCAGGAAATAGAAACTTAAGAGAAATACA-3', from the 9-27 gene (bind to STAT 1:2 dimers); SIE, 5'-AGCTTCATTTCCCGTAAATCCCTA-3', from the c-Fos gene (bind to STAT 1 and 3 homo- and heterodimers). The T4 polynucleotide kinase, Klenow kit, and herring sperm DNA were from Promega Corp. (Madison, WI). Gamma 32P-ATP (7000 Ci/mmol) and alpha 32P-dCTP (3000 Ci/mmol) were from ICN (Costa Mesa, CA). Poly deoxy-inosinic-deoxycytidylic acid and ATP were from Boehringer Mannheim Co. (Indianapolis, IN). Centricons (10 000 molecular weight cutoff) were from Amicon Inc. (Beverly, MA).
Cells and Immunocytochemistry
The BEND cells were isolated from a primary cell culture as described by Staggs et al. [23]. The cell line is deposited and characterized by the American Type Culture Collection (ATCC number CRL-2398; ATCC, Manassas, VA). The ATCC describes methodology for subculturing, propagating, freezing, and products produced by the cells. The BEND cells (
1 x 106 cells) were plated in 150- x 25-mm sterile, polystyrene, tissue culture dishes in 40 ml of complete culture medium (40% Hams F-12 [D-valine added at 0.00085%], 40% valine-modified MEM, 1% antibiotic-antimycotic solution, insulin [0.2 U/ml], 10% fetal calf serum [heat inactivated], 10% horse serum [heat inactivated]) and grown to 90% confluency at 37°C under a humidified atmosphere containing 5% CO2. Changes in cell morphology were observed by light microscopy and photographed.
To confirm the epithelial nature of BEND cells (i.e., presence of cytokeratin markers), epifluorescence studies were conducted as described in Johnson and coworkers [24]. The BEND cells on passage 14 were grown on LabTek four-well chamber slides. Cells were fixed in -20°C methanol for 10 min and rinsed in 0.02 M PBS containing 0.3% Tween-20. Slides were then blocked in antibody dilution buffer (AbDB; two parts 0.02 M PBS, 1.0% BSA, 0.3% Tween-20, pH 8, and one part glycerol) containing 5% normal rabbit serum for 1 h at room temperature. Following a quick rinse with PBS, slides were incubated overnight at 4°C with primary antibody (20 µg/ml AbDB). Following three rinses in PBS for 10 min each, immunoreactive proteins were detected by incubating the slides with fluorescein-conjugated secondary antibody (goat anti-rabbit IgG) for 1 h at room temperature and again washed in PBS (3 x 10 min). Slides were overlaid with a coverglass and Prolong antifade mounting reagent and viewed with a Zeiss Photomicroscope III (Zeiss Inc., Thornwood, NY) equipped with a fluorescein filter set.
To characterize further phenotype of BEND cells, whole cell extracts were obtained from BEND cells from passages 8 and 15 (see below). Protein (2.5, 5, or 10 µg) from extracts of each passage was loaded in duplicate 7.5% acrylamide minigels. Proteins were transferred to nitrocellulose membranes and analyzed for immunoreactivity with vimentin (1:2000 final dilution) and cytokeratin (1:2000 final dilution) by immunoblotting.
Cellular Extracts and Experimental Designs
Cells were washed and incubated in serum-free medium for 45 min. Recombinant bIFN- treatments were added subsequently, as described for specific experiments. At the end of bIFN- incubations, culture medium was discarded and cells were rinsed twice in ice-cold PBS containing 1 mM Na3VO4 and 5 mM NaF at 4°C. Cells were washed briefly in 1 ml of ice-cold whole cell extract buffer (50 mM Tris [pH 8.0], 300 mM NaCl, 20 mM NaF, 1 mM Na3VO4, 1 mM Na4P2O7, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 0.5 mM PMSF, 10% v/v glycerol, 0.5% v/v NP-40, and 10 µg/ml each aprotinin, leupeptin, and pepstatin) or hypotonic buffer (12 mM Hepes [pH 7.9], 4 mM Tris [pH 7.9], 0.6 mM EDTA, 10 mM KCl, 5 mM MgCl2, 1 mM Na3VO4, 1 mM Na4P2O7, 1 mM NaF, 0.5 mM PMSF, 0.6 mM DTT, and 10 µg/ml each aprotinin, leupeptin, and pepstatin [25]). Hypotonic buffer was used to obtain cytosolic extracts and nuclear extracts. Cells were scraped from plates in the presence of 1 ml of the appropriate extraction buffer and incubated on ice for 10 min. Subsequently, cells were lysed by aspiration through a 25-gauge needle. To obtain whole cell extracts, lysates were centrifuged (13 000 x g; 2 min) and supernatants used for immunoprecipitation or immunoblotting. To obtain cytosolic extracts and nuclear extracts [25], lysates were centrifuged (13 000 x g; 20 sec). Supernatants (cytosolic fractions) were adjusted to a final concentration of 60 mM KCl and centrifuged for 30 min in a microcentrifuge (13 000 x g) and used for immunoprecipitation. The pellet (from initial centrifugation of lysate) was resuspended in 1 ml hypotonic buffer, layered over 9 ml of a 30% sucrose solution of hypotonic buffer, and centrifuged at 7000 x g for 8 min at 4°C. The pellet (nuclear fraction) was resuspended in 150 µl high salt buffer (hypotonic buffer containing 300 mM KCl and 20% glycerol) and incubated in a rotating device for 30 min and clarified by centrifugation (13 000 x g; 20 min). Resulting supernatant (nuclear extract) was used straight for electrophoretic mobility shift assays or diluted to 1 ml with hypotonic buffer, adjusted to 60 mM KCl and either used for immunoprecipitation (STAT proteins) or concentrated with a centricon and used for immunoblotting (IRF-1).
Dose response of STATs phosphorylation to bIFN-
To test for specificity of bIFN- on STAT phosphorylation and choose an optimal concentration to use in subsequent experiments, seven plates of BEND cells were assigned randomly to receive 0, 3.125, 6.25, 12.5, 25, 50, or 100 ng/ml b-IFN- for 15 min. Whole cell extracts were obtained and immunoprecipitated with anti-STAT-1 and -2 antibodies and analyzed by immunoblotting for anti-phosphotyrosine and for anti-STAT-1 and -2.
Time dependence of STAT phosphorylation
This experiment was designed to study the phosphorylation response of STATs to bIFN- over time. Seven plates of BEND cells were assigned randomly to receive 0 (control) or 50 ng/ml bIFN- for 3, 8, 15, 30, 60, or 120 min. Whole cell extracts were immunoprecipitated with anti-STAT-1 and -2 antibodies and analyzed by immunoblotting for anti-phosphotyrosine and anti-STAT-1 and -2. The same experiment was conducted with the whole cell extracts immunoprecipitated with anti-STAT-3 antibody and analyzed by immunoblotting for anti-phosphotyrosine and anti-STAT-3. These preliminary experiments were conducted to establish a time line for subsequent nuclear translocation experiments.
Nuclear translocation of STATs
The aim of this experiment was to study the time responsiveness of phosphorylation of STATs to bIFN- in cytosolic extracts and nuclear extracts obtained from the same cells. Plates of BEND cells were assigned randomly to receive 0 (control) or 50 ng/ml bIFN- for 1, 3, 8, 15, 30, 60, or 120 min. Cytosolic extracts and nuclear extracts from each plate were immunoprecipitated with anti-STAT-1 and -2 antibodies and analyzed by immunoblotting for anti-phosphotyrosine and anti-STAT-1 and -2. The same experiment was conducted, but cytosolic extracts and nuclear extracts were immunoprecipitated with anti-STAT-3 antibody and analyzed by immunoblotting for anti-phosphotyrosine and anti-STAT-3. To confirm that the time-responsiveness to bIFN- was specific, this experiment was repeated, except that bIFN- was not added at any time point, and cytosolic extracts and nuclear extracts were immunoprecipitated with anti-STAT-1, -2, and -3 antibodies simultaneously (1 µg each).
Epifluorescence studies were conducted to further verify bIFN--induced STAT nuclear translocation. After growing on chamber slides, BEND cells were serum starved for 24 h and either left untreated or incubated with bIFN- (50 ng/ml) for 30 min. Cells were fixed, incubated with anti-phospho-STAT-1, anti-STAT-2, or anti-phospho-STAT-3 antibodies, and visualized by fluorescence microscopy as described earlier.
Coimmunoprecipitation of STATs
To verify whether bIFN- induces formation of complexes of STATs 1 and 2 with STAT-3, eight plates of BEND cells were assigned randomly to be treated with 0 (control) or 50 ng/ml of bIFN- for 1, 3, 8, 15, 30, 60, or 120 min. Cytosolic extracts and nuclear extracts were obtained from each plate and immunoprecipitated with anti-STAT-3. Immunoprecipitated proteins were examined by immunoblotting analysis for STAT-1 and -2, phosphotyrosine, and STAT-3, in that order. In a separate experiment designed to test whether bIFN- induces formation of complexes of STAT-3 with STATs 1 and 2, cytosolic extracts and nuclear extracts were immunoprecipitated with STATs 1 and 2 and proteins were analyzed by immunoblotting for STAT-3, phosphotyrosine and STATs 1 and 2, in that sequence.
Time response of IRF-1 in response to bIFN-
To verify whether bIFN- was able to induce synthesis of IRF-1, nuclear extracts were obtained from one plate immediately after incubation with serum-free medium and from the remaining eight plates after they received 0 or 50 ng/ml bIFN- for 1, 2, 4, or 6 h. Abundance of IRF-1 in nuclear extracts was measured by immunoblotting and densitometry.
Immunoprecipitation and Immunoblotting
Swollen protein-A agarose beads (40 µl slurry in extract buffer/sample) were incubated with anti-STAT-1 and anti-STAT-2 or with anti-STAT-3 antibodies (1 µg each/sample) for 2 h at 4°C in a rotating device. Beads were then washed thrice in either whole cell extract buffer (for whole cell extracts) or hypotonic buffer (for cytosolic extracts and nuclear extracts). Samples were immunoprecipitated overnight at 4°C, washed three times in 1 ml of the appropriate extraction buffer, and once with 62.5 mM Tris-HCl, pH 6.8. Proteins were solubilized by incubating beads for 5 min at 100°C with reducing Laemmli buffer [26]. Samples were pulse-spun in a microcentrifuge and supernatants used for immunoblots.
Immunoprecipitated proteins or proteins in nuclear extracts were separated in 7.5% acrylamide minigels by one-dimensional SDS-PAGE and transferred to nitrocellulose membranes. For phosphotyrosine blots, membranes were blocked in 2% (w/v) gelatin in Tris-buffered saline and Tween 20 (TBST; 10 mM Tris, 150 mM NaCl, pH 7.6 supplemented with 0.1% v/v Tween-20), and for detection of STATs and IRF-1, membranes were blocked in 5% (w/v) nonfat dried milk in TBST for 2 h. Membranes were washed in TBST and incubated with the following antibodies as appropriate: anti-phosphotyrosine monoclonal (PY-20; 1:1000 in 2% [w/v] gelatin in TBS; 1 h), anti-STAT-1 (1:1000) and anti-STAT-2 (1:1000) polyclonal, anti-STAT-3 (1:1000), or IRF-1 polyclonal (1:10 000) for 2 h in 5% nonfat dried milk diluted in TBS. Membranes were washed in TBST, incubated for 1 h in a 1:8000 dilution of horseradish peroxidase (HRP)-linked anti-mouse IgG in TBST containing 2% (w/v) gelatin (phosphotyrosine blots), or anti-rabbit IgG in TBST containing 5% nonfat dried milk (STATs and IRF-1 blots). Membranes were washed with TBST and proteins detected by ECL following the manufacturer's instructions. To measure abundance of STATs in membranes previously probed for phosphotyrosine, membranes were stripped, neutralized [27], and incubated with anti-STAT antibodies as described above.
Electrophoretic Mobility Shift Assays
Probes Annealed, complementary oligodeoxyribonucleotides corresponding to ISRE and to SIE were labeled radioactively for experiments. Probes for ISRE were prepared by end labeling. Two hundred nanograms of double-stranded ISRE DNA were mixed with 5 µl of 10x polynucleotide kinase buffer, 40 µl of water, 2 µl of gamma 32P-ATP (333 µCi), and 1 µl T4 polynucleotide kinase in a microfuge tube (500 µl) and incubated for 30 min at 37°C. Then, 5 µl of 250 mM ATP were added to the reaction mixture and incubated for 10 min at 37°C. Finally, 10 µl of tRNA (1 mg/ml in water) and 5 µl of phenol saturated in 10 mM Tris, 1 mM EDTA, pH 8 buffer (TE) were added. Radiolabeled probe was purified using Sephadex-G10 chromatography followed by precipitation with ethanol. Specific activity of labeled probe was 210 cpm/ng. Radiolabeled probe was stored at -20°C until use in electrophoretic mobility shift assays.
Probes for SIE were prepared by fill-in. Double-stranded SIE DNA (200 ng) was mixed with 4 µl of 10x Klenow polymerase buffer, 16 µl of water, 6 µl of dNTP mix (3.33 mM each dATP, dTTP, and dGTP), 10 µl of 32P-dCTP (100 µCi), and 2 µl of Klenow and incubated for 30 min at 37°C. The remainder of the procedure was as described for labeling of the ISRE probe. Specific activity of labeled probe was 63 cpm/ng.
Experiments Samples included nuclear extract (10 µg, experiment 1; 7 µg, experiment 2), 8 µl of 5x electrophoretic mobility shift assay buffer (10 mM Tris-HCl, pH 7.4, 60 mM KCl, 1 mM MgCl2, 0.5 mM EDTA, 0.5 mM DTT, 10% glycerol, 4 µg herring sperm DNA), 2 µl poly-deoxyinosinic-deoxycytidylic acid, and sterile water to a final volume of 40 µl. Incubations could also contain a 10-fold excess of unlabeled probe, specific antibodies (1 µl each of anti STAT-1, anti-STAT-2, or anti-IRF-1, or 10 µl of anti-STAT-3) or rabbit IgG (1 µl of a 200-µg/ml solution). Mixtures were incubated at 37°C for 30 min. Labeled probes (ISRE, experiment 1; SIE, experiment 2) were added to the reaction mix (100 000 cpm) and incubated at 37°C for an additional 30 min. Five microliters of loading buffer (40% glycerol, 60% water, 0.01% xylene cyanol) were added to each tube and entire samples loaded onto 6% acrylamide, pre-electrophoresed gels. Gels were run for 5 to 6 h at a constant 30 mA. Dried gels were exposed to x-ray films overnight at -80°C.
Experiment 1 was designed to identify the nature of proteins contained in complexes induced by bIFN- and bound to an ISRE-labeled probe, using antibodies against STATs 1, 2, 3, IRF-1, and rabbit IgG. Experiment 2 was designed to identify the nature of proteins contained in complexes induced by bIFN- and bound to an SIE-labeled probe, using antibodies against STATs 1, 3, and rabbit IgG.
Densitometry and Statistical Analyses Abundance and phosphorylation of STATs in the dose response, nuclear translocation, and coimmunoprecipitation experiments were analyzed by densitometry (AlphaImager 2000, Alpha Innotech Corporation, San Leandro, CA). Density values for bands were adjusted for the background of each individual lane.
Densitometric values were analyzed by least squares analysis of variance, using the GLM procedure of SAS [28]. Independent variables were dose and replicate for dose response experiments and time and replicate for time response experiments. For time response experiments, means were compared using a series of preplanned orthogonal contrasts (0 versus the average of 3, 8, 15, 30, 60, and 120 min; 3 versus the average of 8, 15, 30, 60, and 120 min; 8 versus the average of 15, 30, 60, and 120 min; 15 versus the average of 30, 60, and 120 min; 30 versus the average of 60 and 120 min; 60 versus 120 min). To analyze abundance of IRF-1, independent variables were gel, treatment, time, gel by treatment, gel by time, and treatment by time. Because there was only a negligible signal for IRF-1 in the absence of bIFN-, data were reanalyzed for the effect of time on samples treated with bIFN- only. Quantitative differences were considered significant with P < 0.10 (range P, 0.06 to 0.10). Investigators accepted a 10% chance of committing a type 1 error. Maximal densitometric values were evident visually among doses of IFN- and among times for the various responses. Furthermore, phosphorylation of STAT proteins and subsequent localization to the nucleus were verified independently with immunocytochemical-epifluorescence experiments that examined either phosphorylated STAT proteins (STAT-1 and STAT-3) or abundance (STAT-2) in the nucleus.
RESULTS
Morphological Characterization of BEND Cells
After evaluation of BEND cells at passage 11 by light microscopy, it is suggested that cells are largely of epithelial origin, because they present a cobblestone morphology (Fig. 1a). Elongated cells (perhaps of stromal origin) and cells with a larger, round, and flattened appearence were present in fewer numbers than cells with a cobblestone appearence. There was a noticeable change in cell morphology with cells of a more advanced passage containing a greater proportion of cells with a larger, round, and flattened appearence (passage 33; Fig. 1b). Only cells between passages 8 and 25 were used for experiments. There were no noticeable changes in response to bIFN- regarding phosphorylation of STATs for any of the passages tested.
|
Epifluorescence studies indicated that BEND cells stained positive for both cytokeratin and vimentin, but staining for desmin was much less (Fig. 1, c–g).
Strong immunoreactivity for cytokeratin was noted both for passages 8 and 15 (Fig. 1h). Staining for vimentin was always less intense than for cytokeratin and was not observed in extracts from cells of passage 8. No increase in vimentin intensity was observed after passage 15 (data not shown).
Validation of Immunoprecipitation and Immunoblotting Procedures
Immunoblots of STAT proteins immunoprecipitated with anti-STAT antibodies show a clear enrichment of the appropriate STATs (STAT-1, 91 kDa; STAT-2, 113 kDa; STAT-3, 90 kDa; Fig. 2a). Moreover, proteins were not detected in the absence of immunoprecipitating antibodies. Only nonspecific bands were present when samples were immunoprecipitated with normal rabbit serum (i.e., same banding pattern noted regardless of antibody used). Similarly, STATs were present in immunoblots probed with anti-phosphotyrosine and anti-STAT antibodies but were absent when immunoblots were probed without antibody or with normal rabbit serum (Fig. 2, b and c). In the absence of bIFN-, there was no change in phosphorylation of STATs 1, 2, or 3 over time (data not shown).
|
Effects of bIFN- on Tyrosine Phoshorylation of STAT Proteins
Phosphorylation of STATs 1 and 2 increased with as little as 3.125 ng/ml of bIFN- for STAT-2 and 6.25 ng/ml for STAT-1 (Fig. 3). There was a gradual increase in phosphorylation of STAT-1 that reached a plateau between 6.2 and 100 ng/ml. Phosphorylation of STAT-2 did not change significantly with doses of bIFN- greater than 3.125 ng/ml. Abundances of STATs 1 and 2 were unchanged across all doses. A dose of 50 ng/ml was chosen for all subsequent experiments.
|
Bovine IFN- induced tyrosine phosphorylation of both STATs 1 and 2 (Fig. 4a) and STAT-3 (Fig. 4c) after a 3-min treatment. Maximum phosphorylation occurred at 8 min for STATs 1 and 2 but only at 15 min for STAT-3. For STATs 1, 2, and 3, phosphorylation gradually decreased to reach control levels by 60 min. Exposure to bIFN- did not affect abundance of STATs 1, 2 (Fig. 4b), and 3 (Fig. 4d).
|
Effects of bIFN- on Nuclear Translocation of STAT Proteins
STAT-1
Abundance of STAT-1 in cytosolic extracts remained constant regardless of the length of time of exposure to bIFN- (Fig. 5a). In contrast the abundance of STAT-1 in nuclear extracts increased by 8 min in response to bIFN-. The STAT-1 levels reached a maximum at 15 min, remained elevated, and slightly decreased by 120 min. Bovine IFN- stimulated a transient increase in phosphorylation of STAT-1 in the cytosol that was first observed after 3 min, maintained maximum levels between 8 and 15 min, and decreased to control levels by 120 min (Fig. 5b). However, phosphorylation was delayed in the nucleus, and maximum phosphorylation only occurred after a 15- to 30-min exposure to bIFN-. Phosphorylation decreased by 60 min.
|
STAT-2
In the cytosol there were no changes in the abundance of STAT-2 after exposure to bIFN- (Fig. 5c). Abundance of STAT-2 in nuclear extracts increased after an 8-min exposure to bIFN- and reached a maximum at 15 min, remained elevated up to 30 min, and then decreased to control levels. Phosphorylation of STAT-2 in the cytosol increased after 3 min, remained elevated between 3 and 15 min, and gradually decreased to control levels in all other time points (Fig. 5d). Phosphorylation of STAT-2 in the nucleus was detected in only one of three experiments, where it increased by 15 min and remained elevated for up to 120 min.
STAT-3
In the nucleus, bIFN- increased abundance of STAT-3, which lasted from 8 to 30 min (Fig. 5e). Tyrosine phosphorylation of STAT-3 increased to reach maximum levels at 15 min in the cytosol (Fig. 5f) and at 30 min in the nucleus. Levels of phosphorylation decreased abruptly after 60 and 120 min exposure to bIFN-.
In epifluorescence studies, 30-min treatments with bIFN- also increased the abundance of phosphorylated STATs 1 and 3 and abundance of total STAT-2 in the nucleus of BEND cells (Fig. 6).
|
Effects of bIFN- on Dimerization of STAT Proteins
Coimmunoprecipitation of STAT-1 with STAT-3
In cytosolic extracts, abundance of STAT-1 complexed with STAT-3 decreased during the 30- and 60-min samples but increased again by 120 min (Fig. 7a). There was an increase in abundance of STAT-1 associated with STAT-3 in the nucleus after 8 min of exposure to bIFN-. The STAT 1:3 complex remained elevated up to 30 min and then declined.
|
Coimmunoprecipitation of STAT-2 with STAT-3
In the cytosol, abundance of STAT-2 decreased to reach its minimum levels between 8 and 60 min but increased to control levels at 120 min (Fig. 7b). In nuclear extracts, bIFN- stimulated a time-dependent association of STAT-2 with STAT-3 that increased after 8 min to reach a maximum at 15 min and decline thereafter.
Coimmunoprecipitation of STAT-3 with STATs 1 and 2 There was a decrease in the abundance of STAT-3 in the cytosol at 15 min. Abundance remained similar to control levels in all other time points (Fig. 7c). In the nucleus, STAT-3 association with STATs 1 and/or 2 increased to reach maximum levels at 8 min and then decreased by 60 min.
Effects of bIFN- on DNA Binding of STAT Proteins
Binding to the ISRE
Experiment 1 revealed that no specific complexes were formed in the absence of nuclear extracts (Fig. 8, lane 1) or when nuclear extracts obtained from untreated cells were used (lane 2). The apparent s3 complex present in lane 1 is an artifact and was not observed in experiment 2 (Fig. 9, lane 1). Formation of the s1 complex was noticed when nuclear extracts from cells treated with bIFN- were used (lane 3), and the intensity of such complex was reduced when a 10-fold excess cold ISRE probe was added to the reaction mixture (lane 4). Addition of an anti-STAT-1 antibody caused a reduction in intensity of s1 that may be associated with the formation of the low-intensity s2 complex. Anti-STAT-2 antibody abolished formation of the s1 complex but stimulated formation of an s3 complex that was not completely resolved in the gel. Formation of an anti-STAT-2-induced s3 complex was confirmed in experiment 2 (Fig. 9, lane 3). Anti-STAT-3, anti-IRF-1, and rabbit IgG had no effect on migration or intensity of the s1 complex.
|
|
Binding to the SIE
In experiment 2, no specific complexes were formed in the absence of nuclear extracts (Fig. 9, lane 4) nor when nuclear extracts from untreated cells were used (lane 5). Treatment with bIFN- induced formation of a major and a minor complex (s4 and s5, respectively; lane 6). Formations of both s4 and s5 complexes were abolished completely when a 10-fold excess of unlabeled SIE probe was added (lane 7). Addition of antibodies against STATs 1 and 3 revealed that s4 was composed of three distinct subcomplexes (Fig. 9b), a faster (s4c), an intermediate (s4b), and a slower (s4a) migrating complex. Presence of anti-STAT-1 antibody displaced both s4b and s4c, but intensity of s4a was actually enhanced. Preincubation with anti-STAT-3 caused displacement of s4a and s4b to form the supershifted complex s6. In contrast, s4c migration remained unaltered. Because s4b was displaced both by anti-STAT-1 and -3 antibodies, it is suggested that the subcomplex contains both STATs and is probably a STAT 1:3 heterodimer. Also, because s4c was displaced when anti-STAT-1 antibody was used, but not anti-STAT-3, it is suggested that s4c represents a STAT 1:1 homodimer. Similarly, s4a displacement with anti-STAT-3 but not anti-STAT-1 indicates that it is composed of STAT 3:3 homodimers. The s5 was formed of two subcomplexes, a slower and a faster migrating. The faster migrating subcomplex was shifted in the presence of anti-STAT-1 antibody, whereas anti-STAT-3 antibody shifted the slower migrating subcomplex. Addition of rabbit IgG did not alter the intensity or mobility of s4.
Effects of bIFN- on Induction of IRF-1
There was a significant time by treatment interaction on abundance of IRF-1 (Fig. 10; P < 0.03). Treatment with bIFN-, but not with medium alone, stimulated synthesis of IRF-1 in a time-dependent manner (P < 0.04). The IRF-1 increased after 1 h of treatment to reach a maximum at 2 h and gradually decreased after 4 and 6 h of exposure to bIFN-.
|
DISCUSSION
Presence and activation of tyrosine phosphorylation of STAT proteins by bIFN- has been shown previously in nonreproductive tissues (MDBK cells [27]) as well as bovine uterus [9–11]. The present series of experiments demonstrate that bIFN- can stimulate the JAK-STAT pathway of signal transduction in BEND cells. The STAT proteins 1, 2, and 3 are present, and after exposure to bIFN-, they become tyrosine-phosphorylated in a dose- and time-dependent manner and associate to form complexes. Such complexes migrate to the nucleus of cells and bind to specific response elements (i.e., ISRE and SIE) present in the regulatory region of genes stimulated by interferons. Moreover, synthesis of the transcription factor IRF-1 was demonstrated. These findings support the concept that the JAK-STAT signal transduction system is indeed functional in BEND cells.
Morphology of BEND cells as well as positive staining for cytokeratin in epifluorescence and immunoblotting studies indicated that they are indeed epithelial. Positive staining for vimentin could be interpreted to mean that BEND cells were contaminated by stromal cells. However, while vimentin is commonly used as a general marker of cells originating from the mesenchyme, vimentin is frequently coexpressed with other members of the intermediate filament family such as cytokeratins, in cultured normal and neoplastic cells [29]. Indeed, Johnson and others [24] recently reported coexpression of both cytokeratin and vimentin in lines of ovine luminal and glandular epithelial endometrial cells. Moreover, vimentin was not detected in immunoblots from BEND cells of an early passage (passage 8). Perhaps expression of vimentin is associated with aging of BEND cells in culture.
Similar to endometrial explants [30–32], BEND cells secreted both interferon-stimulated gene 17 (ISG17) and granulocyte chemotactic protein-2 (GCP-2) in response to bIFN- [11, 23]. Therefore, it is suggested that they are an adequate model for studying IFN-stimulated signal transduction. Due to the relative low abundance of STAT proteins in cells, the immunoprecipitation approach was chosen to concentrate STATs and facilitate detection and analysis of STATs in immunoblots. The anti-human STAT antibodies specifically immunoprecipitated and detected bands for bovine STATs 1, 2, and 3, based on their predicted molecular weight on immunoblots. Therefore, all experiments were conducted using anti-human STAT antibodies.
The initial objective of the present study was to determine a dosage of bIFN- to be used in the experiments. Based on the calculated Kd for the ovine IFN- receptor in endometrium, the physiological dosage for bIFN- would be between 0.2 and 11.6 ng/ml [6]. Phosphorylation of STAT-1 and STAT-2 increased in BEND cells to bIFN- doses in the 3.1 to 12.5 ng/ml range, which is considerably lower than what has been used in other experiments [7, 33]. Present results indicate that bIFN- is able to stimulate tyrosine phosphorylation of STAT-1, but not STAT-2, in a dose-dependent manner. Perhaps if doses between 0 and 3.125 ng/ml of bIFN- had been used, a more gradual, dose-dependent tyrosine phosphorylation of STAT-2 also would have been detected. The-50 ng/ml dose was considered adequate for the subsequent studies conducted. This dose effectively suppressed basal and oxytocin-induced PGF2 secretion in primary culture of endometrial epithelial cells [6] and also suppressed phorbol ester-induced production of PGF2 in BEND cells [34].
As in typical signal transduction paradigms mediated through tyrosine phosphorylation, bIFN- stimulates transient phosphorylation of STAT proteins in whole cell extracts [35]. The dynamics of phosphorylation indicate the presence of a pool of STAT proteins readily available for phosphorylation. However, after a period of maximum phosphorylation (between 8 and 15 min), STATs become refractory to further bIFN--induced tyrosine phosphorylation, which declines. Decreased phosphorylation of STATs after exposure to bIFN- longer than 15 min may reflect degradation of STATs [36] or dephosphorylation of STATs through the actions of phosphatases [37, 38]. Because abundance of STATs is unchanged over time, the latter hypothesis appears most applicable. This agrees with Haspel and coworkers [39], who followed 35S-labeled STAT-1 throughout an IFN-
treatment cycle of 4 h. Only about 10% of STATs were degraded, although 20% to 30% of STATs were in the nucleus by 20 min of treatment. Recent reports have identified tyrosine phosphatases from the families of suppressor of cytokine signaling (SOCS) and protein inhibitor of activated STAT (PIAS) molecules as negative regulators of the JAK-STAT pathway [40–42].
Generally, bIFN- induced an increase in the abundance of STAT proteins in the nucleus over time, which reached a plateau and then decreased. Except for numerical differences for STAT-2, this occurred without reciprocal changes in abundance of STATs in cytosolic extracts. This could indicate that only a small fraction of STATs present in the cytosol actually migrated to the nucleus and the immunoblot technique used was not sensitive enough to detect a decrease of STATs in the cytosolic extracts. Other reports support the concept that STAT molecules cycle to the nucleus as tyrosine-phosphorylated molecules, reaching a maximum at 20 to 30 min, and later return quantitatively to the cytoplasm as nonphosphorylated molecules [39]
Phosphorylation of STAT-1 increased and peaked sooner in cytosolic extracts than in nuclear extracts. Therefore, it is suggested that STAT-1 is initially phosphorylated in the cytoplasm and then translocates to the nuclear compartment. After maximum phosphorylation of STAT-1 at 30 min in the nucleus there was a steep decrease in phosphorylation observed at 60 min. During the same time frame, abundance of STAT-1 in the nucleus decreased proportionally less. This indicates that dephosphorylation of STAT-1 occurs while the protein is still in the nucleus. This concept is supported by the work of David and others [37] who identified a nuclear tyrosine phosphatase that deactivated interferon-regulated STATs and down-regulated transcription of interferon-activated genes.
Phosphorylation of nuclear STAT-2 was only detected directly in one of three immunoblotting experiments. However, an increase in abundance of STAT-2 in the nucleus in response to treatment with bIFN- was verified both by immunoblotting and epifluorescence studies. Because nuclear translocation is usually associated with tyrosine phosphorylation of STATs, it is proposed that although observed only once, tyrosine phosphorylation of STAT-2 in fact occurred in the present study.
In contrast to STATs 1 and 2, changes in phosphorylation of STAT-3 in the cytosolic extracts and nuclear extracts were almost parallel, indicating a rapid translocation of phosphorylated STAT-3 to the nuclear compartment. Similar to STAT-1, there is an abrupt decrease in phosphorylation between 30 and 60 min in both the cytosolic extracts and nuclear extracts, but such a decrease is more accentuated in the cytosolic extracts. Because this decrease is not followed by a concomitant change in abundance of STAT-3, it indicates a faster removal of phosphorylated STAT-3 in the cytosolic extracts compared to the nuclear extracts. Epifluorescence studies further indicated nuclear translocation of STATs 1, 2, and 3 in response to bIFN-.
In the coimmunoprecipitation experiments, there was an overall trend of reciprocal changes in abundance of STATs in nuclear extracts compared to cytosolic extracts. Therefore, an increase in abundance of STATs in nuclear extracts was usually paralleled by a decrease in the cytosolic extracts, and abundance in the cytosolic extracts returned to control levels by 120 min. This indicates that homo- and heterodimers of STATs were formed in the cytoplasm, migrated to the nucleus, and reaccumulated in the cytoplasm.
Association of STATs was noted even without exposure to bIFN-, both in nuclear extracts and cytosolic extracts, indicating a bIFN--independent basal level of association of STATs. In support of our findings, Stancato and coworkers [43] showed that STAT heterodimers (1:2 and 1;rc3) exist in the cytosol prior to cytokine stimulation. There was a faster rate of increase in abundance of STAT-2 in the nucleus compared to STAT-1, associated with STAT-3. This is in conflict with data from Darnell [13] who reported a weak binding of STAT dimers 2:3, compared to 1:2 or 1;rc3. However, Ghislain and Fish [25] reported formation of a STAT 2:3 complex after stimulation of U266 cells with interferon-. Moreover, after maximum abundance at 15 min, there was a sharp decrease of STAT-2 in nuclear extracts, while STAT-1 was still associated with STAT-3 in large amounts by 30 min. It is proposed that the dynamics of association with STAT-3 is different for STATs 1 and 2. Furthermore, due to the apparently greater abundance of STAT-1 in cells compared to STAT-2, one should expect greater formation of STAT 1:3 complexes, rather that 2:3. Because the opposite is the case, bIFN- apparently preferentially induced formation of STAT 2:3 complexes.
Data from electrophoretic mobility shift assays indicated clearly the formation of a specific complex (s1) when nuclear extracts from bIFN--treated cells were used. The radiolabeled ISRE probe is from the 9-27 interferon-induced gene and contains the classical ISRE sequence of nucleotides [12, 44]. Therefore, complex s1 probably is comprised of the ISGF-3 proteins STAT-1, STAT-2, and p48 [12, 45] that are known to bind to such a sequence. Indeed, antibody for STAT-2 caused complete displacement of s1. This could be due to two reasons. First, binding of STAT-2 antibody to STAT-2 on the ISGF-3 complex may cause conformation changes that prevent appropriate binding of ISGF-3 to the ISRE probe. A second possibility is that a supershifted complex was formed but became too bulky to be resolved in a 6% acrylamide gel. Other reports demonstrating that supershifted ISGF-3 complexes used lower percentage acrylamide gels (4.5% [25]; 5% [45]). This preferred alternative is substantiated by the formation of an s3 complex. Lack of alteration of s1 by anti-STAT-3 antibody or rabbit IgG further confirmed specificity of the ISRE probe to the ISGF-3 complex that does not contain STAT-3. Absence of a supershift in response to anti-STAT-3 also was noted by Yang and coauthors in Daudi cells [45]. Transcription of the 9-27 interferon-inducible gene can be stimulated by IRF-1 through binding at the ISRE of that gene [16]. This was not noted in the present experiment, because anti-IRF-1 caused no supershift. This probably occurred because nuclear extracts used in the present study were from cells treated with bIFN- for only 30 min, which is not enough time to allow synthesis of sufficient amounts of IRF-1 to bind to ISRE (based on Fig. 10, maximal protein expression of IRF-1 was not noted until 2 h exposure to bIFN-).
In experiment 2, bIFN- induced formation of two specific complexes, s4 and s5 (SIE probe). The SIE sequence is present in the c-Fos gene and shown to be regulated by binding of STAT-1 [46] and STAT-3 [47] homo- and heterodimers. Indeed, Yang and coauthors [48] reported binding of STAT-1 and -3, but not STAT-2 to SIE-labeled probes after incubations with nuclear extracts from interferon-treated cells. Preincubation with anti-STAT-1 and anti-STAT-3 antibodies yielded formation of distinct supershifted complexes, indicating that both proteins are present and have the ability to bind specific DNA sequences. The supershifted complexes were similar to those reported by Yang and others [48]. Preincubation with anti-STAT-1 antibody caused supershift of the faster migrating s4c subcomplex, but the supershifted subcomplex could not be visualized in the gel. Examination of data in Yang and others [48] shows that the supershifted subcomplex elicited by preincubation with STAT-1 antibody is slower migrating than the supershifted subcomplex formed by preincubation with STAT-3 antibody. Because the supershifted subcomplex elicited by STAT-3 was located at the very top of the electrophoretic mobility shift assay gel, it is presumed that the STAT-1 antibody supershifted subcomplex was of higher molecular weight and therefore did not enter the gel. Alternatively, STAT-1 antibody may have interfered with binding of STAT-1-containing complexes (i.e., STAT 1:1 and STAT 1:3) to the SIE probe. This would explain why s4a was more intense when anti-STAT-1 was added to the preincubation mixture (lane 8) compared to lane 6. Because binding of STAT complexes to the labeled SIE is probably competitive, the s4a complex, probably formed by STAT 3:3 homodimer, had a greater chance to bind to the SIE, causing an increase in intensity of that subcomplex.
Collectively, results from the electrophoretic mobility shift assay experiments support the concept that bIFN- stimulates synthesis of proteins through the JAK-STAT pathway. The working hypothesis is that bIFN--induced proteins inhibit the pathway of PGF2 production in the endometrium. In fact, one such protein is the ISG17, a bovine ubiquitin cross-reactive protein [11], whose promoter region contains five putative ISRE elements. For example, ISG17 forms conjugates with cytosolic proteins, possibly altering their cellular activity [32]. Conjugated proteins may include proteins involved in the synthesis of PGF2, such as the oxytocin receptor, estrogen receptor, protein kinase C, phospholipase A2, and/or cyclooxygenase 2.
Another way to influence negatively the PGF2 generation pathway is to stimulate synthesis of transcription repressors. In the ewe, IFN- induces synthesis of IRF-2, a transcription repressor [18]. Moreover, the promoter region of the estrogen receptor gene, which is involved in luteolysis, contains binding sites for IRFs (IRE [22]). Furthermore, promoter deletion experiments indicated that treatment with bIFN- decreased expression of an estrogen receptor-luciferase reporter construct, but such a decrease was noted only if the region containing the IRE was present [22]. In cattle, the oxytocin receptor probably also is involved in the luteolytic cascade. The oxytocin receptor gene has been sequenced recently and also contains an IRE element that binds both IRF-1 and IRF-2 [21]. Nuclear extracts from endometrium of pregnant cows can bind to the IRE in mobility shift assays. However, binding cannot be supershifted with anti-IRF-1 or -2 antibodies, indicating that other members of the IRF family may be involved [21].
In the present study, bIFN- induced IRF-1 protein expression in BEND cells from passage 17. Attempts were made to measure induction of IRF-2 after exposures to bIFN- ranging from 1 to 48 h, but failed probably due to use of heterologous reagents (anti-human IRF-2 antibody; data not shown). However, because IRF-2 expression can be induced by IRF-1 [16], it is possible that such a response takes place in bIFN--stimulated BEND cells. Alternatively, IRF-1 could regulate transcription of unrecognized genes with potential effects on the PGF2 generation cascade.
ACKNOWLEDGMENTS
The authors thank Dr. R. Michael Roberts for providing the recombinant bIFN- and Dr. Douglas W. Leaman for providing the ISRE and SIE oligodeoxyribonucleotides.
FOOTNOTES
1 This work was supported partially by grant 98-35203-6367 from the NRI Competitive Grants Program/USDA, Florida Agricultural Experimental Station Journal Series no. R-07779, and by NIH grant NIH HD 32475-06. 
2 Correspondence: William W. Thatcher, Dept. of Animal Sciences, Shealy Dr. and Ritchie Rd., Gainesville, FL 32611. FAX: 352 392 5595; thatcher{at}animal.ufl.edu
3 Current address: Centro de Biotecnologia em Reprodução Animal, Faculdade de Medicina Veterinária e Zootecnia, Universidade de São Paulo, Pirassununga, SP 13630-000, Brazil.
4 Current address: Institute de Reproducción Animal, Facultad de Ciencias Veterinarias, Universidad Central de Venezuela, Maracay, Aragua 2101, Venezuela.
Accepted: October 2, 2000.
Received: May 17, 2000.
REFERENCES
of cows in response to recombinant interferon-. J Dairy Sci 1995; 78:1921–1931.[Abstract] regulate basal and oxytocin-induced secretion of prostaglandins F2 and E2 by epithelial cells and stromal cells in the endometrium. Reprod Fertil Dev 1994; 6:193–202.[CrossRef][Medline] in bovine endometrial cells. Biol Reprod 1999; 60:656–663. and interferon- interact with the same receptors in the bovine endometrium. J Biol Chem 1994; 269:13544–13550. and is responsible for their disparate cytotoxicities. Proc Natl Acad Sci U S A 1995; 92:12270–12274.). Endocrine 1997; 6:31–37.[Medline] modulates phorbol ester-induced production of prostaglandin and expression of cyclooxygenase-2 and phospholipase-A2 from bovine enndometrial (BEND) cells. Biol Reprod 2000; 63:417–424.-activated STAT1 by the ubiquitin-proteasome pathway. Science 1996; 273:1717–1719. -mediated cell signalling. J Biol Chem 1995; 270:25709–25714.R1 chain of the human type I interferon receptor. J Biol Chem 1996; 271:8057–8061.This article has been cited by other articles:
|
|
 |
|
  M. M. Joyce, J. R. Burghardt, R. C. Burghardt, R. N. Hooper, L. A. Jaeger, T. E. Spencer, F. W. Bazer, and G. A. Johnson Pig Conceptuses Increase Uterine Interferon-Regulatory Factor 1 (IRF1), but Restrict Expression to Stroma Through Estrogen-Induced IRF2 in Luminal Epithelium Biol Reprod, August 1, 2007; 77(2): 292 - 302. [Abstract] [Full Text] [PDF]
|
|
|
0.1; experiment was replicated twice)
