| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Embryo |
USM 503 MNHN,3 UMR 8646 CNRS-MNHN, U565 INSERM, Département Régulation, Développement et Diversité Moléculaire, Muséum National d'Histoire Naturelle, 75005 Paris, France
INSERM U365,4 Institut Curie, 75248 Paris, France
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(IFN) as well as IFN
are unable to trigger massive nuclear translocation of STAT1 in these cells, even at high cytokine concentrations. Conversely, nuclear accumulation of STAT1 was readily observed in murine L929 somatic cells under the same conditions. However, in the absence of any stimulation, both tyrosine (Y701p) and serine (S727p) phosphorylated forms of STAT1 were already detected in the nuclei of oocytes and early embryos. Phosphorylated STAT1 appeared concentrated in large nuclear dots, which were identified by indirect immunofluorescence and electron microscopy as clusters of interchromatin granules (IGCs or speckles). A similar distribution was also observed for the serine (S727p) phosphorylated form of STAT3 as well as for tyrosine (Y689p) phosphorylated STAT2. Western blot analysis confirmed that STAT factors present in mouse oocytes are predominantly phosphorylated. In parallel, we showed that the transcription of two IFN-target genes, namely interferon regulatory factor-1 (IRF-1) and suppressor of cytokine signaling-1 (SOCS-1) is indeed increased in two-cell embryos in response to IFN. Altogether, our results suggest that, despite the lack of massive nuclear accumulation of STAT1 in response to exogenous IFNs and the permanent presence of phosphorylated STATs in the nucleus, JAK/ STAT pathways are functional during early development.
cytokines, early development, embryo, signal transducers, signal transduction
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-driven signaling pathway, and most particularly STAT1 and IFN target genes, in fully grown oocytes and during preimplantation stages of mouse embryos.
As several other cytokines and growth factors, IFN triggers intracellular responses through the JAK/STAT pathway, i.e., binding to specific heterodimeric receptors, followed by phosphorylation of Janus kinases (JAKs) [2] leading to the phosphorylation on a single tyrosine residue and homodimerization of the latent cytoplasmic transcription factors of the STAT (signal transducer and activator of transcription) family (reviewed in [3–5]). Binding of IFN to its receptor triggers the phosphorylation of JAK1 and JAK2 that in turn specifically activates STAT1 (by phosphorylation of Y701), as well as, in certain cell types, STAT3 [6, 7], while STAT2 is mainly activated by IFN [8]. STAT dimers are then actively translocated to the nucleus [9–11], although no nuclear localization sequence (NLS) has clearly been characterized on the STAT proteins to date [12, 13].
In addition to tyrosine phosphorylation (Y701) of STAT1, which appears sufficient for dimerization, nuclear translocation and transcription transactivation, STAT1 can also undergo S727-phosphorylation in response to IFN. S727-phosphorylation may be a result of the activation of some MAP kinases (mitogen-activated protein kinases, reviewed in [14]) and has been reported to enhance the transcriptional activity of STAT1 ([15]) although having no influence on its DNA binding capacity ([16]). Finally, although STAT1 appears to permanently shuttle between cytoplasm and nucleoplasm apart from any stimulation and/ or phosphorylation ([17]), significantly higher levels of Y701p- and S727p-STAT1, as well as massive but transient nuclear accumulation of STAT1 could only be observed upon stimulation of cells with IFN ([18]).
By interacting with specific DNA sequences, STAT1 drives the transcription of IFN-inducible genes [19]. IRF-1 (interferon regulatory factor-1) has been characterized as an immediate target gene transcribed in response to IFN [20] and is involved in cellular responses to IFN (reviewed in [21]). As many cytokines, IFN also enhances the transcription of SOCS (suppressor of cytokine signaling) genes [22], which encode for proteins that negatively regulate the JAK/STAT pathways by interacting with JAKs (reviewed in [23]). Indeed, JAK/STAT pathways are evolutionarily conserved and knock-out of genes encoding the different proteins involved in these cascades have demonstrated their pivotal role during development in Drosophila as well as in mammals (reviewed in [24]). Given the importance of the JAK/STAT pathway, they are subjected to tight negative regulatory mechanisms aiming to limit and/ or terminate the signaling events (reviewed in [25]).
We report here that, while STAT1 accumulates rapidly in the nuclei of L929 cells upon stimulation with either IFN or IFN, nuclear accumulation does not occur in mouse oocytes and preimplantation embryos even in the presence of higher doses of the cytokines and despite the presence of STAT1 mRNA and protein. Surprisingly, however, phosphorylated forms of STAT1 were detected in the nuclei of oocytes and early embryos, independently of IFN stimulation. They accumulate in discrete subnuclear structures subsequently identified as interchromatin granules (IGC, also termed speckles). Similarly, phosphorylated forms of STAT2 and STAT3 were also detected in nonstimulated oocytes with the same nuclear pattern. Despite this particular situation, transcription of IRF-1 and SOCS-1 genes, which are two IFN early target genes, is enhanced in two-cell embryos in response to the cytokine, indicating that the IFN signaling pathway is functional during preimplantation development.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Normal mice were of the C57Bl6/CBA strain. Sv129 mice, harboring an invalidated IFNGR1 gene [26], were a generous gift from Kader Thiam (Institut Pasteur, Lille, France). Their care and housing were in accordance with the institutional guidelines of the French Ministerial Ethics Committee (Ministère de l'Agriculture et de la Pêche, Direction de la Santé et de la protection Animale, Paris, France) and under the supervision of authorized investigators in accordance with local ethical regulations. Ovarian oocytes at the germinal vesicle (GV) stage were collected from 4- to 8-week-old mice, abundantly rinsed and kept in M2 medium [27] supplemented with dibutyryl cyclic AMP (dbAMPc, 100 µg/ml; Sigma, France) as described in [28]. Zona pellucida of mouse oocytes and early embryos was removed by an acidic Tyrode solution, pH 2.5 [29].
One-cell embryos were recovered from 6- to 8-week-old C57Bl6/CBA females stimulated to superovulate and mated with males of the same strain. Embryos were cultured up to the blastocyst stage in M16 medium (Sigma) without fetal calf serum (FCS; Gibco BRL, France) [28]. Hatching blastocysts were occasionally placed on 0.1% gelatin-coated coverslips in DMEM (Gibco BRL) supplemented with 20% FCS to obtain adherent, expanded blastocysts that were not cultured more than 8 days post-hCG.
To inhibit RNA polymerase II (pol II) activity, -amanitin (20 µM) was added to M2 + dbAMPc medium. Oocytes were fixed after 0, 2, 4, and 6 h of treatment and their transcriptional status was analyzed after BrUTP microinjection [30].
Cell Culture
Murine L929 fibroblasts were cultured on coverslips at 37°C under 5% CO2 in DMEM medium supplemented with L-glutamine and 10% FCS. Cells were used at subconfluence.
Stimulation of Cells and Mouse Oocytes and Preimplantation Embryos by Interferons
Murine L929 cells and hatched blastocysts were starved for 30 min by incubation in fresh medium devoid of FCS. Oocytes, preimplantation embryos, and starved L929 cells were stimulated with recombinant murine (rMu-) IFN (R&D, 5000–50 000 U/ml) or rMu-IFNA (PBL; Biomedical Laboratories, 2000–20 000 U/ml) for 0–60 min at 37°C under 5% CO2. For oocytes, early embryos and L929 cells stimulation, cytokines were diluted in fresh medium, i.e., M2, M16, and DMEM, respectively. After stimulation, cells (or oocytes or embryos) were rapidly rinsed three times with PBS before being processed for indirect immunofluorescence.
Indirect Immunofluorescence
Cells, oocytes, and embryos were rinsed in phosphate buffer saline (PBS), fixed with 4% paraformaldehyde (PFA) in PBS for 30 min at room temperature and permeabilized with 0.2% Triton X100 in PBS for 30 min at room temperature before being processed for indirect immunofluorescence. This procedure was already shown to allow for the penetration of antibodies to the nucleus [30, 31]. They were postfixed in 2% PFA for 20 min at room temperature. Chromatin was stained with Hoechst 33342 (2 µg/ml) or Sytox Green (250–500 mM). For double-labeling (e.g., Y701p-STAT1 and S727p-STAT1 or Y701p-STAT1 and Sm protein), the procedure was the same except that the two primary and the two secondary antibodies were mixed. Slides were mounted with Citifluor (Citifluor Company, UK). Each experiment was performed on 10–30 oocytes or embryos at a given developmental stage (n: number of cells) and was repeated at least three times (e: number of experiences). All experiments included a control set (about 10 oocytes or preimplantation embryos) incubated without primary antibody.
Antibodies
Primary antibodies used for indirect immunofluorescence (IF) and Western blotting (WB) were a polyclonal rabbit anti-mouse STAT1 (p91) (dilutions IF 1:50, WB 1:1000; Santa Cruz Biotechnology); a polyclonal rabbit IgG raised against the S727-phosphorylated form of mouse STAT1 (dilutions IF 1:100, WB 1:1000; Upstate Biotechnology); a polyclonal goat IgG raised against the Y701-phosphorylated form of mouse STAT1 (dilutions IF 1:50, WB 1:500; Santa Cruz Biotechnology); the Y12 mouse monoclonal antibody against Sm proteins, a generous gift of Dr. J.A. Steiz, Yale University, New Haven, CT (dilution IF 1:1000); a polyclonal rabbit IgG anti-S727-phosphorylated STAT3 (dilutions IF 1:50, WB 1:1000; Upstate Biotechnology); a polyclonal rabbit IgG anti-Y705-phosphorylated STAT3 (dilutions IF 1:100, WB 1:1000; Upstate Biotechnology); a polyclonal rabbit IgG anti-STAT3 (dilutions IF 1:100, WB 1:1000; Santa Cruz Biotechnology). These antibodies were already used for Western blot analysis on somatic cells by Sancéau et al. [32]. A rabbit immunoaffinity-purified IgG anti-Y689-phosphorylated STAT2 (dilutions IF 1:100, WB 1: 1000; Upstate Biotechnology) and a rabbit polyclonal anti-STAT2 (dilutions IF 1:100, WB 1:1000; Upstate Biotechnology), the latter two were a generous gift from Dr. S. Pellegrini, Institut Pasteur, France. The non-cross-reactivity of the pairs of antibodies used for codetection experiments was evaluated and controls without primary antibody were systematically performed in parallel for IF experiments.
Secondary antibodies used in IF experiments were a fluorescein (FITC)- or rhodamine-red-conjugated donkey anti-rabbit IgG (H+L) (dilution 1:100 and 1:5000, respectively; Jackson ImmunoResearch Laboratories); a rhodamine-red-conjugated donkey anti-goat IgG (H+L) (dilution 1:2000; Jackson ImmunoResearch Laboratories); a FITC- or Texas red-conjugated donkey anti-mouse IgG (H+L) (dilution 1:100 and 1:200, respectively; Jackson ImmunoResearch Laboratories).
Image Capture and Analysis
Conventional microscopy was performed with a Zeiss inverted microscope (Axiovert 35), equipped with a cooled CCD camera (Photometrics Type KAF 1400, 12-bit dynamic range) coupled to IpLab Spectrum imaging software. Confocal microscopy was performed with an inverted Nikon microscope equipped with Bio-Rad Laser-Sharp MRC-1024 confocal laser scanning software. Quantification of the fluorescence was performed on confocal pictures.
Measures of the nuclear to cytoplasmic fluorescence intensity ratio were performed for each cell using IpLab Spectrum imaging software. Intensity was measured within a rectangular frame of identical size located successively over the nucleus and cytoplasm image. Three independent measures were made for each cell and the mean value of the intensity ratio treated using kaleidagraph software. Significance of the differences was evaluated by Student test.
Electron Microscopy and Immunogold Labeling
Oocytes and embryos were fixed, embedded in Unicryl resin (BB International), and processed for immunogold labeling as already described [30]. Primary antibodies were diluted 10 times less than for IF. Secondary antibodies were a goat anti-mouse IgG (H+L) antibody or a rabbit anti-goat antibody conjugated with 10-nm gold particles (dilution 1:50; TEBU, France).
Western Blots
About 400 ovarian oocytes at GV stage were recovered in a minimum volume of PBS containing an antiprotease cocktail (1 mM PMSF, 50 µg/ml aprotinine, and 50 µg/ml leupeptin) with or without phosphatase inhibitors. L929 cell lysates were always prepared with the antiprotease cocktail and phosphatase inhibitors. SDS-PAGE on 8% acrylamide gel and semidried transfert onto nitrocellulose membrane were performed as usual. Secondary antibody was a horseradish peroxidase (HRP)-conjugated goat anti-rabbit (dilution 1:50 000; Sigma), or rabbit anti-goat (dilution 1:50 000; Sigma). Membranes were intensively washed and stripped with Restore Western Blot Stripping Buffer (Pierce Perbio Science, Bezons, France) before being reprobed with primary antibody (i.e., anti-S727p-, anti-Y701p-, anti-STAT1, anti-S727p-, anti-Y705p-, and anti-STAT3). The signal was detected by means of enhanced chemiluminescence (Super Signal Pierce, Perbio Science, France). Molecular mass of detected proteins was evaluated by comparison with full-range rainbow molecular weight markers (Amersham Biosciences, France).
The mRNA Extraction and RT-PCR
The mRNA extraction procedure from a definite number (around 100) of oocytes or embryos at different stages has been described [1]. Reverse transcription (RT) for murine STAT1 was carried out on a 10-embryo equivalent [1]. Polymerase chain reaction (PCR) was performed on a five-embryo equivalent (10 µl of RT products), using 0.25 µM antisense primer 5'-gTgTTCTgAATATTTCCCTCCTgggCC-3' and 0.25 µM sense primer 5'-gATgTCTCgTTTgCg ACCATCCg-3' [33] in a final volume of 50 µl. The amplification program was a first denaturation step at 94°C for 5 min, followed by 35 cycles at 94°C for 1 min, 65°C for 1 min, 72°C for 1 min, and a final extension step at 72°C for 10 min.
PCR amplification of IRF-1 cDNA obtained by reverse transcription was performed on a five-embryo equivalent (10 µl of RT products). The same procedure as above was used except that the final primers concentration was 0.1 µM. Primers for muIRF-1 were antisense 5'-CTTCATCTCCgTgAAGACATg-3' and sense 5'-CCTgATgACCACAgCAgTTAC-3' [34]. The amplification program was a first denaturation step at 94°C for 5 min, followed by 35 cycles at 94°C for 1 min, 65°C for 1 min, 72°C for 1.5 min, and a final extension step at 72°C for 10 min. Similar PCR conditions were used to amplify SOCS1 cDNA with antisense 5'-gCAgCTCgAAAAggCAgTCg-3' and sense 5'-CACCTTCTTggTgCgCgACA-3' primers [35] for 35 cycles (94°C for 1 min, 58°C for 1 min, and 72°C for 1 min).
The PCR products were separated by electrophoresis on a 2% agarose gel containing 0.25 µg/ml ethidium bromide (BET) and then observed with an ultraviolet imager (Appligene, Inc.).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We first investigated if STAT1 was expressed in mouse oocytes and preimplantation embryos, both at mRNA and protein levels. RT-PCR on whole RNA extracts revealed a single fragment at the 258-base pair (bp) size expected for the STAT1 cDNA, abundant in mouse oocytes and one-cell embryos, less abundant in two-cell embryos, to reach a virtually undetectable level in four-cell embryos. STAT1 mRNA was then again clearly detected from the morula stage on, as well as in oocytes from IFNGR1–/– mice (Fig. 1A). STAT1 protein was clearly detected in the cytoplasm at all stages analyzed from oocyte to hatched blastocyst by indirect immunofluorescence (Fig. 1B, first line, and data not shown). In addition, in IFNGR1–/– oocytes (data not shown), STAT1 labeling appeared similar to that observed in normal oocytes. Control without primary antibody confirmed the specificity of the observed labeling (Fig. 1B, third line).
|
STAT1 Does Not Accumulate in Nuclei of Mouse Oocytes and Preimplantation Embryos in Response to IFNs Stimulation
Nuclear translocation of STAT1 upon activation of the IFN-driven signaling pathway is already amply documented. However, no nuclear accumulation of STAT1 in mouse oocytes could be observed after 5–60 min of stimulation by either IFN (5000–20 000 U/ml) or IFN (2000–50 000 U/ml) (Fig. 1B, second line). Similarly, massive nuclear translocation was not observed in response to IFNs in mouse preimplantation embryos from one-cell to hatched blastocyst stage (data not shown). Stimulation of IFNGR1–/– oocytes with IFN did not lead to nuclear accumulation of STAT1 (data not shown). By contrast, murine L929 cells stimulated under the same conditions with the lowest concentrations of IFN (5000 U/ml) or IFN (2000 U/ml) showed a progressive nuclear accumulation of STAT1 (Fig. 2A), which peaked after 20 min (Fig. 2B). A similar time course of nuclear translocation was observed using a primary antibody specific for the tyrosine (Y701) phosphorylated form of STAT1 (data not shown). Control without primary antibody showed no labeling in L929 cells (Fig. 2C).
|
Phosphorylated Forms of Various STATs Are Accumulated and Colocalized in the Nuclei of Unstimulated Mouse Oocytes and Preimplantation Embryos
Because stimulation of mouse oocytes and preimplantation embryos with IFN or IFN failed to induce nuclear accumulation of STAT1, it appeared possible that STAT1 did not undergo phosphorylation in response to IFNs. To test this hypothesis, mouse oocytes and early embryos were stimulated with IFN 5000 U/ml from 5 to 30 min and IF was performed using a primary antibody specifically directed against Y701p-STAT1. Y701p-STAT1 was indeed detected in the cytoplasm and the nucleus of oocytes, whatever the incubation time (data not shown), but, also more surprisingly, in oocytes of the control group not exposed to IFN (Fig. 3A). Similar cytoplasmic and nuclear distributions of Y701p-STAT1 were also observed in nonstimulated preimplantation embryos from the late one-cell to the hatched blastocyst stage (Fig. 3B and data not shown). The nuclear distribution of Y701p-STAT was particularly intriguing because it always appeared accumulated in nuclear round substructures or dots. This labeling was not observed in the absence of primary antibody (Fig. 3C).
|
Moreover, in the case of oocytes, this labeling differed between the two categories found among fully grown oocytes, SN (surrounded nucleoli) and NSN (nonsurrounded nucleoli) [28]: numerous (around 25) nuclear dots of relatively heterogeneous size (diameter ca. 2 µm) in NSN oocytes, and a diffuse labeling with sometimes 0–5 large dots (diameter ca. 4 µm) in SN oocytes (Fig. 4A). In addition, the distribution pattern in NSN oocytes was changed to that of SN oocytes when polymerase II-dependent transcription was inhibited (Fig. 4D). A similar punctuated pattern of Y701p-STAT1 was observed in embryos of all stages starting from late one-cell to hatched blastocyst (Fig. 4C and data not shown).
|
When using a specific antibody against the S727-phosphorylated form of STAT1 (S727p-STAT1), a similar nuclear labeling pattern was obtained in oocytes and embryos at all developmental stages (data not shown) and double-labeling experiments revealed a perfect colocalization of the two phosphorylated forms of STAT1 (Fig. 4B). Colocalization was also observed when double labeling was sequentially performed (i.e., S727p- and Y701p-STAT1 or Y701p- and S727p-STAT1). Finally, an identical distribution of both Y701p- and S727p-STAT1 was observed in nuclei of oocytes from IFNGR1–/– mice (data not shown).
Because of the particular nuclear distribution of Y701- and S727-phosphorylated forms of STAT1 in mouse oocytes and preimplantation embryos, we wondered whether it could be a common feature of other STAT family members, for example, STAT3, which has also been reported to be activated and translocated to the nucleus in response to IFN [6, 7]. Indeed, while STAT3 was mostly cytoplasmic in nonstimulated mouse oocytes, S727p-STAT3 was also detected in nuclear dots, the number of which depended on the oocyte type, and appeared colocalized with Y701p-STAT1 (data not shown). Although we failed to detect Y705p-STAT3 under our indirect immunofluorescence conditions, Western blot results clearly indicate that it is present in unstimulated oocytes (see Fig. 6) but its precise location remains to be determined.
|
Finally, STAT2, which is usually phosphorylated in response to type I IFNs, was also detected in the cytoplasm of unstimulated oocytes and a specific antibody for Y689p-STAT2 localized the latter form in nuclear dots similar in size and number to those observed for phosphorylated forms of STAT1 (data not shown).
Phosphorylated Forms of STATs Are Associated with Interchromatin Granule Clusters
Because of the intriguing presence of phosphorylated STAT factors in the nuclei of unstimulated oocytes, we investigated whether the dots represent already characterized subnuclear compartments. Double immunostaining for Y701p-STAT1 and Sm proteins, which are components of the nuclear structures recognized as interchromatin granules by electron microscopy and speckles by fluorescence microscopy, showed that both antibodies label the same nuclear structures (Fig. 5A).
|
Immunogold detection of Y701p-STAT1 in nonstimulated mouse oocytes showed that gold particles were abundant at the plasma membrane, particularly at the level of the microvilli (data not shown), and were associated with cytoplasmic fibrillar structures (Fig. 5B, 1), as well as with nuclear pore complexes (Fig. 5B, 5). In the nuclei of nonstimulated mouse oocytes, groups of gold particles were specifically associated with clusters of interchromatin granules (IGCs) (Fig. 5B, 1–3) but not with other nuclear structures, such as the nucleolus (Fig. 5B, 2) or twisted fibrils (Fig. 5B, 4). Importantly, a similar nuclear and specific labeling of interchromatin granules was observed using the anti-STAT1 antibody, although weaker than with the anti-Y701p-STAT1 antibody (Figs. 5B, 6). Moreover, no labeling was observed in the absence of the primary antibody (anti-STAT1 or anti-Y701p-STAT1 antibody, data not shown).
Western Blots Confirm the Presence of Phosphorylated STATs in Unstimulated Mouse Oocytes
The presence of Y701p- and S727p-STAT1 (Fig. 6A, left panel) as well as of S727p- and Y705p-STAT3 (Fig. 6B, left panel) in unstimulated mouse oocytes was confirmed by Western blots. Moreover, comparison of the intensity of the bands suggested, at least in the case of STAT3, that Y705p-STAT3 represents an important proportion of the overall STAT3 pool. Similar results were obtained in the presence (data not shown) or absence (Fig. 6) of phosphatase inhibitors in oocytes samples. When performed on whole lysates of nonstimulated L929 cells, Western blot revealed low basal levels of Y701p-STAT1 and S727p-STAT1, whereas both phosphorylated forms of STAT1 were enhanced in IFN-stimulated cells (Fig. 6A, right panel; compare NS and S). However, neither Y705p- nor S727p-STAT3 appeared to be enhanced upon IFN stimulation in L929 cells (Fig. 6B, right panel; compare NS and S). Reprobing the membranes with anti-STAT1 or anti-STAT3 antibodies confirmed that phosphorylated proteins detected were STAT1 and STAT3, respectively. In addition, no signal was detected all over the membranes in the absence of primary antibody.
Western blot on oocyte lysates with an anti-STAT2 antibody gave rise to a couple of immunoreactive bands, the upper one being also recognized by an anti-Y689p-STAT2 antibody. Surprisingly, however, these two bands exhibited apparent molecular masses of about 72 and 81 kDa, respectively, whereas the expected STAT2's molecular mass of 113 kDa was observed on Western blot of nonstimulated L929 cells (data not shown).
IFN Enhances the Transcription of Early Target Genes in Mouse Two-Cell Embryos
Because of the particular distribution of both latent and phosphorylated STAT1 in mouse oocytes and early embryos and the apparent lack of nuclear accumulation of STAT1 upon IFN stimulation, we next investigated whether IFN was able to induce the transcription of IFNs' target genes. We chose to explore the expression of IRF-1 mRNA in mouse oocytes and preimplantation embryos because the IRF-1 gene has been characterized as an early IFN-target gene. As shown in Fig. 7A, IRF-1 mRNA was present in nonstimulated oocytes from C57/CBA mice, as well as in mouse one-cell embryos, but its level appeared to decrease from the two-cell stage on, to become undetectable in four-cell and morula stages. A very low level of IRF-1 mRNA was again detectable in blastocysts and hatched blastocysts. In addition, when two-cell embryos (stage at which the major burst of zygotic transcription occurs) were stimulated by IFN (5000 U/ml) for 2 h, the level of IRF-1 mRNA appeared similar to that observed in unstimulated two-cell embryos but was slightly increased after 6 h of stimulation (Fig. 7A; compare lanes 3, 8, and 9).
|
Cytokines also rapidly induce the expression of SOCS gene encoding for a family of negative regulators of the JAK/STAT pathways, particularly SOCS-1 transcription has been reported to be upregulated in a STAT1-dependent manner in response to IFN [22]. We thus investigated whether SOCS-1 expression was enhanced upon IFN stimulation in mouse oocytes and early embryos. As shown in Figure 7B, SOCS-1 mRNA was already detected in unstimulated mouse oocytes and one-cell, but not in two- and four-cell, embryos, and again at the morula, blastocyst and hatched-blastocyst stages. Upon stimulation of two-cell embryos with 5000 U/ml of IFN, SOCS-1 mRNA was unambiguously expressed as early as after 2 h of stimulation (Fig. 7B; compare lanes 3, 8, and 9).
Because PCR was applied on a definite and constant number of embryo equivalents and the volume of PCR products analyzed was constant, comparison of the intensity of the bands observed for embryos of the same developmental stage suggests that IFN signaling pathway may be functional and able to enhance transcription of IFN-inducible genes in mouse preimplantation embryos.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
signaling pathway during early development. A very intriguing result was the detection of both Y701- and S727-phosphorylated forms of STAT1 (reviewed in [14]) in the nucleus of unstimulated oocytes and preimplantation embryos when using specific primary antibodies. This observation was confirmed by Western blot on mouse oocytes, unambiguously demonstrating that both forms of phosphorylated STAT1 are permanently present and quite stable, as they were detected even in the absence of phosphatase inhibitors during the sample preparation.
Moreover, the phosphorylated forms of STAT1 were highly accumulated in specific nuclear substructures of oocytes and preimplantation embryos. Several lines of evidence suggest that these structures represent the speckles/ clusters of interchromatin granule: i) the Y12 anti-Sm protein antibody labels the same structures; ii) their distribution and size in NSN and SN oocytes follow those already described for the speckles in oocytes [31]; iii) they increase in size and decrease in number upon inhibition of pol II-dependent transcriptional activity, as already described for speckles in somatic cells [36]. The association of Y701p-STAT1 with interchromatin granules was further confirmed at the electron microscopy level. IGCs were also specifically, although more weakly, labeled by an anti-STAT1 antibody on electron microscopy sections. The fact that a very low labeling was observed by indirect immunofluorescence using the same antibody after in toto labeling suggests that nuclear STAT1 is less accessible to this antibody except from the surface of sections. This may be because of changes in STAT1 conformation and/or accessibility or yet because of posttranslational modifications of phosphorylated STATs, i.e., ubiquitination [37] or small ubiquitin-related modifier conjugation [38, 39]. In this regard, experimental results suggest that a pool of Y701p-STAT1 could be ubiquitinated in unstimulated mouse oocytes (our unpublished results).
The detection of Y701p-STAT1 by electron microscopy in unstimulated oocytes at the plasma membrane, at the level of cisternae of the endoplasmic reticulum and of nuclear pore complexes suggests a permanent trafficking of STAT1 molecules, phosphorylated at the plasma membrane and then targeted to the nucleus. Although we cannot exclude the possibility that, upon stimulation, a small pool of cytoplasmic STAT1 is activated and translocated to the nucleus, in a proportion too low to be detected with our methods, the emerging picture is that STAT1 is not massively translocated into the nucleus upon IFN stimulation, but that phosphorylated STAT1 already resides in the nucleus. Indeed, although not quantitative, WB data suggest that, in the case of STAT3, phosphorylated forms represent an important proportion of the cellular pool of this transcription factor. Thus, the fact that a pool of STAT1 is already tyrosine and/or serine phosphorylated may in part explain the lack of massive nuclear accumulation of STAT1 upon IFN stimulation. Because i) mouse oocytes are freshly collected and ii) phosphorylated forms of other STATs are detected by IF, it can be hypothesized that different JAK/STAT pathways are functional, possibly triggered by various cytokines and growth factors produced by ovarian cells and/or oocytes themselves and acting in an autocrine/paracrine manner. This hypothesis is quite supported by our results showing the detection of both Y701p- and S727p-STAT1 in the nucleus of oocytes from IFNGR1–/– mice, indicating that, even if involved, IFN signaling is dispensable for the activation of STAT1 in mouse oocytes. In the same line, the detection of Y689p-STAT2 in mouse oocytes strongly suggests that signaling by type I IFNs is functional. Moreover, the expression of IFN mRNA in mouse oocytes and early embryos has already been described [40], suggesting a possible paracrine/autocrine action of IFN on oocytes and early embryos. A possible explanation for the unusual MW we obtained for STAT2 could be the expression of a development-specific spliced variant, as suggested by Darnell [5].
As mouse embryos appear to be relatively independent of the environment for their growth, given that they develop in vitro without addition of any growth factors, cytokines, or serum, the permanent activation of STATs may also be maintained during the preimplantation period by cytokines and growth factors secreted by the embryos themselves, which also express their corresponding receptors [41]. In this respect, the targeted disruption of the STAT3 [42], JAK1 [43], and JAK2 [44, 45] genes leads to embryonic or perinatal lethality, pointing to an important role of the JAK/ STAT pathways during mammalian development.
In addition, the enhanced expression of IRF-1 and SOCS-1 mRNAs in response to IFN in mouse two-cell embryos gives rise to the intriguing hypothesis that the JAK/STAT pathway may be functional in mouse preimplantation embryos but that transcription of IFN target genes does not require massive nuclear accumulation of STAT1. However, our results do not rule out the possibility that parallel signaling pathways emerging from the IFN/ IFN receptor complex may converge to induce IFN-target genes transcription. Nevertheless, the viability of mice with targeted disruption of IFN [46], IFNGR2 [47], or STAT1 [48, 49] genes suggests that IFN signaling is dispensable for development. Our results also show for the first time that SOCS-1 gene is naturally expressed during mouse early development. The possibility that different SOCS genes may be expressed in preimplantation embryos might be of particular importance for the regulation of various JAK/ STAT pathways given the embryonic or perinatal lethality observed for SOCS-3–/– and SOCS-1–/– mice, respectively (reviewed in [23]). Altogether, these data suggest that, although not essential, IFN signaling is functional during the development and may exert some effects on developing embryos at least by indirectly modulating various JAK/STAT pathways by enhancing SOCS-1 expression.
The permanent activation and the lack of massive nuclear accumulation of STATs upon extracellular stimulation in mouse oocytes closely resemble the situation described in some tumoral and transformed cells (reviewed in [50]). Because oocytes and early embryos of different mammalian species i) are able to develop in vitro independently of exogenous factors, ii) express and secrete certain growth factors and cytokines as well as their cognate receptors (reviewed in [51]), and iii) are nonadhesive dividing cells, becoming invasive at the time of implantation, it is interesting to note that, as in transformed cells, STAT1 and STAT3 are found permanently activated by phosphorylation on both tyrosine and serine. Due to their respective roles in the control of the cell cycle and apoptosis in transformed cells, these two STAT proteins may be of particular importance during early development.
Clusters of interchromatin granules are sites of transit/ storage/recycling of several components of the mRNA synthesis and maturation machinery (reviewed in [52]). To our knowledge, association of activated STATs with interchromatin granule clusters has not been described in another cell type. Moreover, the association of Y701p- and/or S727p-STAT1, Y689p-STAT2, and S727p-STAT3 with this subnuclear compartment suggests that this may be a general feature of STAT family members in mouse oocytes and preimplantation embryos. This particular nuclear distribution during early development may reflect the convergence of multiple signaling pathways on STAT transcription factors. By analogy to the situation described for some splicing factors and RNA polymerase II (reviewed in [53]), accumulation of phosphorylated STATs in association with IGCs may reflect the storage of ready-to-serve (i.e., tyrosine- and/or serine-phosphorylated) transcription factors in the neighborhood of transcription sites because the triggering of the transcription is tightly regulated temporally and spatially during early development.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: UMR 8646 CNRS-MNHN/U565 INSERM, Muséum National d'Histoire Naturelle, 53 rue Buffon, 75005 Paris, France. FAX: 01 58 41 50 20; Sandrine.Truchet{at}ibpc.fr
Received: 8 March 2004.
First decision: 2 April 2004.
Accepted: 27 May 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
