Expression, activity, and subcellular localization of the Yin Yang 1 transcription factor in Xenopus oocytes and embryos.

Yin Yang 1 (YY1) is a multifunctional transcription factor that acts as an activator, repressor, or initiator of transcription of numerous cellular and viral genes. Previous studies in tissue culture model systems suggest YY1 plays a role in development and differentiation in multiple cell types, but the biological role of YY1 in vertebrate oocytes and embryos is not well understood. Here we analyzed expression, activity, and subcellular localization profiles of YY1 during Xenopus laevis development. Abundant levels of YY1 mRNA and protein were detected in early stage oocytes and in all subsequent stages of oocyte and embryonic development through to swimming larval stages. The DNA binding activity of YY1 was detected only in early oocytes (stages I and II) and in embryos after the midblastula transition (MBT), which suggested that its potential to modulate gene expression may be specifically repressed in the intervening period of development. Experiments to determine transcriptional activity showed that addition of YY1 recognition sites upstream of the thymidine kinase promoter had no stimulatory or repressive effect on basal transcription in oocytes and post-MBT embryos. Although the apparent transcriptional inactivity of YY1 in oocytes could be explained by the absence of DNA binding activity at this stage of development, the lack of transcriptional activity in post-MBT embryos was not expected given the ability of YY1 to bind its recognition elements. Subsequent Western blot and immunocytochemical analyses showed that YY1 is localized in the cytoplasm in oocytes and in cells of developing embryos well past the MBT. These findings suggest a novel mode of YY1 regulation during early development in which the potential transcriptional function of the maternally expressed factor is repressed by cytoplasmic localization.

Yin Yang 1 (YY1) is a multifunctional transcription factor that acts as an activator, repressor, or initiator of transcription of numerous cellular and viral genes. Previous studies in tissue culture model systems suggest YY1 plays a role in development and differentiation in multiple cell types, but the biological role of YY1 in vertebrate oocytes and embryos is not well understood. Here we analyzed expression, activity, and subcellular localization profiles of YY1 during Xenopus laevis development. Abundant levels of YY1 mRNA and protein were detected in early stage oocytes and in all subsequent stages of oocyte and embryonic development through to swimming larval stages. The DNA binding activity of YY1 was detected only in early oocytes (stages I and II) and in embryos after the midblastula transition (MBT), which suggested that its potential to modulate gene expression may be specifically repressed in the intervening period of development. Experiments to determine transcriptional activity showed that addition of YY1 recognition sites upstream of the thymidine kinase promoter had no stimulatory or repressive effect on basal transcription in oocytes and post-MBT embryos. Although the apparent transcriptional inactivity of YY1 in oocytes could be explained by the absence of DNA binding activity at this stage of development, the lack of transcriptional activity in post-MBT embryos was not expected given the ability of YY1 to bind its recognition elements. Subsequent Western blot and immunocytochemical analyses showed that YY1 is localized in the cytoplasm in oocytes and in cells of developing embryos well past the MBT. These findings suggest a novel mode of YY1 regulation during early development in which the potential transcriptional function of the maternally expressed factor is repressed by cytoplasmic localization.
Yin Yang 1 (YY1 1 ; previously referred to as FIII, NF-E1, ␦, F-ACT1, CF1, or UCRBP) is a GLI-Kruppel family transcrip-tion factor that functions either as a repressor, activator, or initiator of transcription (reviewed in Ref. 1). It is expressed in a wide variety of different cell types and shows a high degree of sequence similarity among divergent organisms (2)(3)(4)(5). The mammalian YY1 protein (414 aa) contains four C-terminal zinc fingers, an N-terminal bipartite transcriptional activation domain, and a number of domains that function to mediate interactions with other proteins (6). The Xenopus YY1 cDNA is highly conserved, but the predicted protein is somewhat smaller at 373 aa (4).
Numerous studies have revealed that YY1 is subject to complex regulatory mechanisms in different cell types. The emerging picture is that the functional diversity of YY1 as an activator, repressor, or initiator of transcription is context-specific and is modulated by interactions with various cellular factors and adaptor proteins. YY1 has been shown to interact physically with a number of proteins including c-Myc (7,8), p300/ CBP (9), CREB (10), ATF/CBP (11), poly(A)DP-ribosyl transferase (12), histone deacetylase (13), the E1A oncoprotein (2), and a number of general transcription factors (14,15).
A variety of tissue culture studies have shown that YY1 controls expression of developmentally regulated genes, and so it is thought to play important roles in development and differentiation (reviewed in Ref. 1). Recognition sites for YY1 have been identified in the regulatory regions of transferring (15), dystrophin (16), and actin (17)(18)(19)(20)(21)(22)(23) genes expressed in differentiating and aging muscle, fetal and embryonic globin genes (24 -27), the immunoglobulin 3 gene expressed in pre-B cells (28), and the pre-T-cell ␣ receptor gene (29). YY1 contributes to the tissue-specific regulation of the prosacrosin gene in testes (30) and to the osteocalcin gene in bone (31). In bone the activity of YY1 appears to be controlled by targeting to the nucleolus and nuclear matrix, (32,33). In developing muscle, proteolytic processing of YY1 is thought to be responsible for cell-specific derepression of sarcomeric ␣-actin genes (23). Despite its apparent association with cell growth and differentiation in diverse cell types, however, very little is known about the expression and regulation of YY1 in vertebrate oocytes and embryos, and the biological role of YY1 during early vertebrate development is poorly understood. A recent study of YY1 in mouse development revealed that it is abundantly expressed in oocytes and early embryos, displays a dynamic pattern of subcellular localization in early development, and is selectively expressed in certain tissues later in development (34). Because heterozygous mutation of YY1 in the mouse resulted in growth and neurulation defects, and full knockout caused embryonic lethality around implantation, it appears likely that YY1 functions at multiple stages of mouse development (34).
The objective of the present study was to identify and characterize YY1 in the Xenopus developmental model system. We found that YY1 mRNA and protein were expressed in early oocytes and were present at relatively constant levels throughout oocyte and embryonic development. The ability of YY1 to interact with its cis-recognition site was restricted to very early oocytes and to post-MBT stages in embryos, suggesting that the DNA binding activity of the maternal factor is modified specifically during development. Expression assays showed that YY1 had no effect on transcription from the thymidine kinase (TK) promoter in oocytes, but we also found a similar lack of transcriptional activity in post-MBT embryos at a point in development when its DNA binding is reactivated. Analysis of the subcellular localization of YY1 showed it is sequestered in the cytoplasm of oocytes and in cells throughout the embryo up to as late as the neurula stage of development. We suggest that cytoplasmic sequestration is a mode of developmental regulation used to inhibit the transcriptional activity of YY1 during early stages of Xenopus development and that YY1 probably functions in cell differentiation events that occur later in development.

EXPERIMENTAL PROCEDURES
Oocyte and Embryo Manipulations-Xenopus laevis frogs were purchased from Xenopus I (Ann Arbor, MI). Oocytes were obtained surgically from adult female frogs, and follicle cells were removed by swirling in calcium-free OR2 (82.5 mM NaCl, 2.5 mM KCl, 1 mM CaCl 2 , 1 mM MgCl 2, 1 mM NaH 2 PO 4 , 5 mM Hepes (pH 7.8) 10 mg/liter streptomycin sulfate, 10 mg/ liter benzyl penicillin) containing 2 mg/ml collagenase (type II, Sigma) for 3 h. Oocytes were washed extensively in OR2 (containing 1 mM CaCl 2 (35) prior to microinjection or preparation of cell extracts. For expression of YY1 in oocytes, 200 pg of in vitro synthesized mRNA was injected directly into the cytoplasm, and oocytes were incubated for 16 h prior to preparation of protein extracts. Expression plasmid encoding the Xenopus YY1 mRNA (4) was a kind gift from E. Beccari, (Universita di Roma La Sapienza, Rome, Italy) Embryos were obtained as described in (36).
Tissue and Cell Extracts-For protein extracts, oocytes or embryos were homogenized in Buffer C (50 mM Tris-Cl, pH 7.9, 20% glycerol, 50 mM KCl, 0.1 mM EDTA, 2 mM dithiothreitol, 10 g/ml each of aprotinin and leupeptin (36), cell debris was pelleted for 5 min at 15,000 ϫ g (4°C), and the supernatants were placed in fresh tubes and frozen in liquid nitrogen. Embryos or stage V and VI oocytes were counted and homogenized in a volume of 10 l/embryo/oocyte. To obtain approximately equal protein concentrations, oocytes at stages III and IV were homogenized in 4 l/oocyte, stages II and III in 2.5 l/oocyte, and stages I and II in 1 l/oocyte. Nuclear and cytoplasmic extracts of embryos were performed as described by Lemaitre et al. (37).
For A6 protein extracts, cells were washed in cold PBS, pelleted, and homogenized by freeze-thawing in a buffer containing 20 mM HEPES, 25% (v/v) glycerol, 0.42 M NaCl, 1.5 mM MgCl 2 , O.2 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, and 0.5 mM dithiothreitol. The homogenate was centrifuged at 10,000 ϫ g in Eppendorf tubes, and supernatants were stored at Ϫ80°C.
Rat liver nuclear extracts were prepared as previously described (38). Rat livers were chopped into small pieces and homogenized in a buffer containing 10 mM HEPES, pH 7.6, 25 mM KCl, 2 M sucrose, 10% v/v glycerol, 1 mM EDTA, 0.15 mM spermine, and 0.5 mM spermidine using a Dounce homogenizer. Homogenates were layered onto sucrose cushions and centrifuged for 20 min at 24,000 rpm in a SW28 rotor. Nuclear pellets were resuspended in nuclear resuspension buffer (10 mM HEPES, pH 7.6, 100 mM KCl, 3 mM MgCl 2 0.1 mM EDTA, 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, 10% glycerol, 1 nM microcystin) supplemented with 0.3 g/ml (NH4) 2 SO 4 , shaken for 30 min, and homogenized with a Dounce homogenizer; they were then centrifuged at 50,000 rpm in a Ti70 rotor. Protein pellets were resuspended in nuclear resuspension buffer and dialyzed to 40 mM KCl.
Nucleic Acid Analyses-Total nucleic acids were extracted from oocytes or embryos as described (39), by homogenization in Buffer A (50 mM Tris-HCl, pH 7.6, 50 mM NaCl, 10 mM EDTA, 0.5% SDS) containing 0.25 mg/ml proteinase K. The homogenate was incubated at 37°C for 1 and extracted with an equal volume of phenol/chloroform, and nucleic acids were recovered by precipitation after the addition of 0.1 volume of5 M ammonium acetate and 3 volumes of ethanol. Ethanol precipitates containing total nucleic acids were resuspended in 400 l of TE (10 mM Tris-HCl (pH 8.0), 1 mM EDTA). To separate DNA from total RNA, 400 l of 8 M LiCl was added, and the mixture was incubated at 4°C overnight. The RNA was pelleted at 15,000 ϫ g for 10 min, resuspended in TE, and stored as ethanol precipitates. DNA was recovered by precipitation after the addition of 3 volumes of ethanol to LiCl supernatants. Northern blots were performed as described previously (40). Blots were prehybridized overnight at 42°C and hybridized overnight at 55°C in hybridization fluid (50% formamide, 0.8 M NaCl, 0.8 M sodium citrate, 10 mM sodium phosphate, 2.5ϫ Denhardts', 250 mg/ml salmon sperm DNA) containing 10 6 cpm probe. The probe was the cDNA insert of the plasmid encoding Xenopus YY1 mRNA (4) isolated by BamHI-HindIII digestion.
Gel Mobility Shift Assays-DNA binding reactions using embryonic or stages V and VI oocyte extracts contained 1 embryo or oocyte equivalent. Binding reactions for early stage oocyte samples contained 2.5 oocyte equivalents of stage III an IV extracts and 10 oocyte equivalents of the batch containing stage I and II oocytes. The ␣ YY1 recognition element probe from the mouse H3.2 gene, (5Ј-ggatcCTCGGCCGTCAT-GGCGCTGCAGGAGGC-3Ј) was end-labeled with [␣-32 P]dCTP with the Klenow fragment of DNA polymerase I. The sequence of AP-1 oligonucleotide probe is (5Ј-CCGGAAAGCATGAGTCAGACAC-3Ј). Each binding reaction contained 0.2 ng of double-stranded labeled probe, 20 g of soluble protein from indicated samples, 0.5 g of poly(dI-dC), 10 mM Tris (pH 7.8), 50 mM NaCl, 1 mM EDTA, 0.5 mM dithiothreitol, and 5% glycerol. Binding reactions were incubated on ice for 20 min and immediately loaded onto 5% nondenaturing polyacrylamide gels containing 6.7 mM Tris-Cl (pH 7.5), 1 mM EDTA, 3.3 mM sodium acetate. Complexes were electrophoresed for 2 h at 150 V, and gels were dried and exposed overnight to x-ray film.
Cell Culture, Stable Transfections, and CAT Assays-A6 cells were grown in L15 medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum (Hyclone) as described previously (41). Chinese hamster ovary (CHO) cells were grown in McCoy's 5A medium supplemented with 10% calf serum, and stable transfections were performed as described previously (5). CAT assays for CHO cells were performed with the CAT enzyme assay system (Promega) using liquid scintillation counting as described in the Promega technical bulletin, part no. TB084. Microinjections and CAT assays for expression in oocytes were performed as described previously (42).
Chromatography-A6 cell extracts (300 l) were applied to 2 ml of heparin-agarose columns (Sigma) at 4°C. Columns were equilibrated to 50 mM KCl with DC50 buffer (20 mM HEPES, pH 7.9, 0.2 mM EDTA, 0.2 mM EGTA, 50 mM KCl, 20% v/v glycerol, 2 mM dithiothreitol, 0.5 mM ZnCl) and proteins were eluted with a stepwise gradient of DC buffer at 50, 250, and 400 mM KCl. At each step, 2 ml of buffer was applied to the column, and 200 l fractions were collected.
Ovary and embryo sections were cut to 8 m, and then sections were placed on glass slides and incubated overnight at 35°C. Sections were de-paraffinated and rehydrated by rapid immersions in a series of xylene/ethanol/PBS solutions (100% xylene, xylene/ethanol (1:1), absolute ethanol, 95% ethanol, 70% ethanol) and then immersed for 3 ϫ 30 s in 1 M PBS and 30 min in 1 M PBS at 4°C. All subsequent steps were carried out at room temperature. Sections were incubated for 30 min in block solution (3% skim milk, 1% Triton X-100, 1 M PBS), and then primary antibodies were applied to slides at a dilution of 1:200 in 1% milk block solution for 1 h. Primary antibodies were rabbit anti-human YY1 polyclonal (Santa Cruz sc-281) and mouse anti-human proliferating cell nuclear antigen (PCNA) monoclonal (Santa Cruz sc-56). Control sections were treated with block solution without primary antibody. Sections were washed (2 ϫ 10 min) in 1 M PBS, incubated in block solution (30 min) and then exposed to secondary antibody (1:200, 30 min). Secondary antibodies for detection of YY1 were biotinylated goat anti-rabbit IgG and for PCNA, biotinylated horse anti-mouse IgG

Identification of YY1 and Its DN Binding Activity in Xeno-
pus-In a previous study, we reported a DNA binding activity in Xenopus embryos specific to the internal coding region activator ␣ element of replication-dependent histone genes, but the proteins in these complexes were not identified (36). Subsequently, a yeast one-hybrid screen of a HeLa cell cDNA library established that the protein interacting with the histone regulatory sequence in mammalian cells was YY1 (5). To confirm the identity of the Xenopus factor, we performed gel mobility shift assays using the ␣ probe, which contains the YY1 consensus sequence, in combination with antibody recognition in vitro using polyclonal anti-YY1 antibodies (Fig. 1A). The addition of YY1 antibodies (sc-281, Santa Cruz Biotechnology) to binding reactions with crude whole cell extracts from the A6 Xenopus kidney epithelial cell line or neurula stage (stage 13) embryos disrupted the specific complex. Similar YY1 complexes and specific antibody disruptions were observed with rat liver nuclear extracts (Fig. 1A). The slower migrating nonspecific complexes were not recognized by YY1 antibodies, and migration of the YY1 complex was not affected in controls using antibodies against other proteins. Thus, it appears that YY1 binds to and retards the mobility of the ␣ probe in Xenopus cell extracts.
The same cell extracts used for mobility shift assays were analyzed on Western blots using YY1 antibody (Fig. 1A, lower  left panel). The amino acid sequence alignment with peptide used for antibody sc-281 is 100% homologous between the human and frog YY1. A protein at 59 kDa was detected in A6 cells and neurula stage embryos (stage 13), and a slightly larger band at 65 kDa was detected in rat liver nuclei. Microinjection of in vitro synthesized mRNA encoding Xenopus YY1 into oocytes resulted in an increased detection of the 59 kDa band (Fig. 1A, lower right panel), evidence that the mammalian antibody actually detected the frog protein in Western blots. The sizes of both the frog and rat YY1 on the blot are larger than predicted from the number of amino acids (41 and 45.5, respectively). Anomalous migration has also been observed for the human protein, which migrates as a 65-68-kDa protein although its predicted size is 44 kDa; this difference is thought to be due to protein structure (1). The protein expressed in both embryos and somatic cells of Xenopus was slightly smaller than mammalian YY1; this is consistent with the relative predicted protein sizes of 373 aa for Xenopus (4) versus 414 aa for rat (2,5).
We next analyzed the YY1 factor by fractionation of Xenopus A6 kidney epithelial cell extracts on heparin-agarose columns. Cell extracts were loaded onto columns and eluted in a stepwise KCl gradient. The majority of specific YY1 element binding activity eluted at 400 mM (Fig. 1B, upper panel). Western blot analysis of column fractions using the YY1 polyclonal antibody showed that the 59-kDa polypeptide appeared most prominently in the fractions containing peak DNA binding activity. The correlation between the appearance of the YY1 signal and specific DNA-binding complexes shown in fractions (Fig. 1B), as well as the supershift of DNA-binding complexes (Fig. 1A), confirm the identity of the YY1 transcription factor in Xenopus extracts. Some YY1 was detected in fractions eluted at 250 mM salt, but there was no corresponding gel shift activity. It is possible that cofactors required for efficient DNA binding were not present in the lower salt fractions, or it could be that the Western blot detection of YY1 protein is much more sensitive than detection of gel shift activity in these assays.
Expression of YY1 during Early Development-Previous descriptions of YY1 as an important factor in cell growth and differentiation prompted us to examine the profile of YY1 mRNA accumulation during oocyte and embryonic development. Northern blots were performed on total RNA isolated from oocytes and embryos at various stages of development ( Fig. 2A). A 4-kilobase mRNA species was detected with the YY1 cDNA probe. A relatively constant amount of YY1 mRNA signal was detected in early and late stage oocytes, unfertilized eggs, in cleavage, blastula and early neurula embryos, and in swimming larvae. Because the zygote does not become transcriptionally active until midblastula stage (43), it appears that transcription of the maternally encoded YY1 gene occurs very early during oogenesis and maternal YY1 transcripts are relatively stable through fertilization and early embryonic development. The YY1 mRNA species detected in Northern blots migrated just below the 28 S ribosomal RNA bands. To rule out the possibility of cross-hybridization to ribosomal RNA, we tested for hybridization of the YY1 probe in poly(A) ϩ and poly(A) Ϫ fractions isolated over oligo(dT) columns. The 4-kilobase YY1 mRNA signal was detected in poly(A) ϩ samples containing mRNA but not in the poly(A) Ϫ sample containing ribosomal RNA (Fig. 2A).
We next characterized the profile of YY1 protein expression during development. In the Western blots shown in Fig. 2B, polyclonal anti-YY1 antibodies were used to detect YY1 in staged whole cell extracts of oocytes and embryos. A single prominent polypeptide migrating at 59 kDa was detected in unfertilized eggs and in developing embryos from early cleavage through to neurula stages. As with the profile of mRNA expression, there were no significant changes in the relative amounts of YY1 protein signal detected throughout these phases of development. We also examined expression of YY1 during oogenesis and through oocyte maturation and fertilization; the Western blot to compare YY1 in oocytes, eggs, and early embryos is shown in Fig. 2C. Similar quantities of the YY1 signal were seen in stage VI oocytes, in unfertilized eggs, and at cleavage stage. However, in early oocytes, YY1 appeared to be significantly smaller at 50 kDa. A more detailed comparison at each stage of oogenesis showed that YY1 was predominantly a 500-kDa protein in early oocytes, but by stage V1 the majority of YY1 migrated at 59 kDa (Fig. 2D).
DNA Binding Activity of YY1 during Oocyte and Embryonic Development-To examine the profile of YY1 DNA binding activity during development, crude whole cell extracts were prepared from staged oocytes, eggs, and embryos and analyzed in gel mobility shift assays using the ␣ probe with the YY1 recognition sequence (Fig. 3A). The specific YY1 complex was detected in stage I and stage II oocytes and in embryos after the MBT, but it was absent in the intervening stages. The addition of YY1 antibodies to DNA-binding reactions disrupted band formation, confirming the presence of YY1 in shifted complexes in these assays. The activity of other transcription factors are known to be stable through this period of development (44), and the DNA binding activity of AP-1 transcription factor is stable in the same extracts (Fig. 3A, lower panel). In addition, the changes in YY1 DNA binding activity through early development were not due to fluctuations in the levels of maternal protein, because the same extracts were used in the Western blots above (Fig. 2) in which relatively constant amounts of YY1 were detected. Also of note is that DNA binding activity was down-regulated in oocytes at stages III-V (Fig. 3A), and so there did not appear to be a correlation between reduced DNA binding activity and the change in molecular mass from 50 to 59 kDa that occurred between stages V and VI.
Previous studies have provided evidence suggesting that the phosphorylation state of YY1 can affect its DNA binding activity (5). We examined whether phosphorylation might affect YY1 during development. Phosphatase treatment of stage VI extracts resulted in a partial recovery of DNA binding (Fig. 3B). This suggests that YY1 binds more efficiently in the hypophosphorylated state, and that the DNA binding activity of YY1 might be differentially regulated by phosphorylation during oocyte development. We also considered the possibility that the increase in the apparent molecular weight of YY1 in late stage oocytes could be due to increased phosphorylation. On Western blots, however, YY1 was detected at 59 kDa in both untreated and phosphatase-treated extracts (Fig. 3B). This suggests the increase in apparent molecular weight of YY1 at stage VI is not attributable to phosphorylation.
Transcriptional Activity of YY1 in Xenopus Embryos-We next examined the transcriptional activity of YY1 using a series of microinjected reporter constructs containing the thymidine kinase (TK) promoter linked to YY1 binding sites. Reporter constructs (Fig. 4A)  elements in forward (YY1-TKCAT) or reverse orientation (YY1R-TKCAT) immediately upstream of the TK promoter, 6 repeats of a mutant YY1 binding site (YY1M-TKCAT), or a single YY1 binding site from the histone H2a coding region activator sequence (CRAS-TKCAT). CAT expression was analyzed in embryos that were allowed to develop to gastrula stage at 4 h past MBT (Fig. 4B). Insertion of YY1 elements did not significantly stimulate or repress activity from the TK promoter in post-MBT embryos. In numerous repeats of this experiment, there were no significant changes in CAT activity relative to controls. In control experiments using CHO cells, we determined that YY1 was able to influence transcription in context of the intact TK promoter (Fig. 4C). YY1 elements stimulated a 3-fold activation over the basal promoter, indicating that these constructs respond to YY1. Therefore, in Xenopus it appears that YY1 is transcriptionally inactive in post-MBT embryos. Similar results were obtained in expression experiments with stage VI oocytes (data not shown), which was expected given the low level of YY1 DNA binding activity at later stages of oocyte development (Fig. 3A). Our findings are in agreement with previously reported experiments showing the YY1 binding site from L1 and L14 ribosomal protein genes did not affect transcription in Xenopus oocytes (45).

YY1 Is Localized in the Cytoplasm of Oocytes and Early
Embryos-It was recently shown that YY1 is localized in the cytoplasm of unfertilized mouse oocytes and displays a mosaic pattern of subcellular localization in E3.5 blastocysts (34). To determine the subcellular location of YY1 during oogenesis, we enucleated oocytes and analyzed nuclear and cytoplasmic fractions for the presence of YY1 by Western blotting. At both stages IV and VI, YY1 was detected in the cytoplasm but was absent in nuclear extracts (Fig. 5A). This finding clearly demonstrates that YY1 is cytoplasmic in late stage oocytes and suggests that the increase in molecular mass from 50 to 59 kDa, which occurs between stage VI and stage VI, is not associated with nuclear relocalization. We also tested for DNA binding activity in both cytoplasmic and nuclear compartments (Fig. 5B). Although YY1 does not normally bind DNA in late stage oocytes, activity was induced by in vitro phosphatase treatment as demonstrated in Fig. 3B. Phosphatase treatment of subcellular portions from stage IV and VI oocytes revealed that YY1 DNA binding could be activated in cytoplasmic extracts but not in nuclear extracts (Fig.  5B). Supershift controls showed that phosphatase-activated complexes were specifically recognized by YY1 antibodies. These experiments confirm the findings of Western blots (Fig.  5A) showing that YY1 is sequestered in the cytoplasm in stage VI and VI oocytes.
The DNA binding activity of YY1 is present in stage I oocytes, switched off, and then reactivated after MBT (Fig. 3A). This pattern prompted us to examine a possible correlation between DNA binding and nuclear localization during development; i.e. DNA binding might be associated with nuclear localization. We hypothesized that YY1 would be a nuclear protein in early oocytes, relocalize to the cytoplasm in later stage oocytes, remain cytoplasmic in the embryo during cleavage and early blastula stages, and then migrate back into nuclei after MBT. To test this hypothesis, we performed an immunocytochemical analysis of developing oocytes and embryos (Fig. 6). We used the same YY1 antibody as described above for Western blots and mobility supershifts. As a positive control we used an anti-PCNA monoclonal antibody, which resulted in nuclear staining in all sections tested. Negative controls were done in which secondary antibodies alone were applied to oocyte and embryonic sections for detection of background staining (data not shown). As a positive control, YY1 antibody was applied to FIG. 4. Transcriptional activity of YY1 in post-MBT embryos. A, diagram of plasmid constructs containing the CAT gene driven by the TK promoter and YY1 elements as described under "Experimental Procedures." B, embryos were microinjected after fertilization with the indicated plasmid constructs and then allowed to develop to gastrula stage, at which point CAT assays were performed (upper panel). The experiment was repeated at least five times with different batches of embryos; a representative CAT assay is shown. CAT activity is expressed as a percentage of the highest relative activity (TK-CAT), which was arbitrarily assigned a value of 100 (lower panel). C, CAT assays of stable transfectant CHO cells. Constructs are indicated above the panel. Activity is presented as percent acetylation. The experiment was performed with triplicate sets of plates. adult Xenopus liver sections, because we had already observed high levels of DNA binding activity in nuclear extracts from rat liver (Fig. 1). A prominent signal was detected in liver nuclei, but YY1 appeared to be localized entirely in the cytoplasm in both early (stage I) and late (stage VI) stage oocytes. A cytoplasmic staining pattern was also observed in cells of developing embryos. YY1 did not appear in the nuclei of cleavage stage embryos, nor in gastrula stage embryos several hours past the MBT, nor in neurula stage embryos at a point in development 12 h past the onset of zygotic transcription.
The finding that YY1 DNA binding activity appears after the MBT (Fig. 3) but appears to remain cytoplasmic (Fig. 6) was surprising and interesting. To investigate this conclusion further and to confirm the results of immunostaining, we tested for the presence of YY1 in biochemically produced nuclear and cytoplasmic fractions from stage 13 neurula embryos. The Western blots shown in Fig. 7 show that YY1 was entirely cytoplasmic. In a control blot using the same fractions, the PCNA antigen was detected only in nuclear and whole cell samples. Thus, it appears that YY1 remains sequestered in the cytoplasm during early embryonic development and does not contribute to early developmental gene expression. DISCUSSION In this study we have demonstrated that YY1 is a maternally derived protein that is present throughout early Xenopus development. YY1 is expressed abundantly in early oocytes and, exists as a cytoplasmic protein in oocytes and early embryos, and its DNA binding activity is restricted to early oocytes and to post-MBT embryos. At first, the developmental profile of DNA binding prompted us to hypothesize that maternal YY1 might function to modulate expression of specific target genes in stage I oocytes and once again after MBT when expression from the zygotic genome commences in the embryo (46). However, we found that YY1 binding sites had no effect on expression from the TK promoter in oocytes or in post-MBT embryos, and so we surmised that endogenous YY1 is transcriptionally inactive during early development. The transcriptional inactivity of YY1 is consistent with our finding that YY1 remains in the cytoplasm of oocytes and in cells throughout the developing embryo well past MBT. Therefore, we suggest that the potential function of YY1 as a transcriptional regulator is specifically repressed by cytoplasmic partitioning during both oogenesis and early embryonic development.
YY1 is a maternally expressed protein in unfertilized oocytes in both mouse (34) and Xenopus. However, the subcellular localization pattern of YY1 in Xenopus is somewhat different from that in mouse development. Donohoe et al. (34) showed that YY1 is cytoplasmic in unfertilized mouse oocytes and becomes nuclear in two-cell embryos coincident with the onset of mouse zygotic transcription, and they suggested that YY1 probably functions to regulate gene expression early in mammalian embryonic development. No such nuclear relocalization occurs in Xenopus embryos at MBT, despite the apparent reactivation of its DNA binding activity at this stage. Unlike mouse embryos in which YY1 becomes nuclear at the 2-cell stage coincident with the onset of zygotic transcription, there is no apparent correlation between nuclear localization and the MBT in Xenopus. Therefore, it would appear that YY1 does not have a direct transcriptional regulatory function at the initial stages of zygotic gene expression in Xenopus embryos or in early development to the neurula stage.
The abundant maternal expression and subcellular localization reported in this study suggest two hypotheses for the potential developmental function of YY1. YY1 may function to regulate the transcription of genes important for differentiation and organ formation, which occur later in development. This has been suggested by numerous studies in various tissue culture model systems (1) and is consistent with the expression pattern of YY1 in mouse embryos and the phenotype of heterozygous YY1 mutants (34). Alternatively, YY1 may play an important role in the cytoplasm during oogenesis and early embryogenesis, perhaps in the transport, storage, or translational regulation of maternal mRNAs. Such a cytoplasmic function would be similar to that proposed for the CCAAT box transcription factor (CBTF), which is retained in the cytoplasm prior to MBT through its RNA binding capacity and is associated with transcriptionally quiescent mRNP complexes (47). Nuclear relocalization of CBTF is associated with degradation FIG. 6. YY1 is a cytoplasmic protein in oocytes and embryos. An immunohistochemical analysis of YY1 and PCNA in oocytes and embryos was done. Oocyte and embryo stages are indicated at the left of each series of panels. In the stage 9 embryo panels, the dorsal part of the embryo is shown. In the neurula panels, the dorsolateral part of the embyro is shown. The bottom panels show YY1 localization in hepatocyte nuclei of adult frog rat liver. of maternal transcripts (47). Unlike CBTF, however, YY1 does not localize to the nucleus after MBT, which would suggest that it might associate with stable mRNAs. Further analyses will be aimed at determining whether YY1 has RNA binding activity and whether association with mRNPs might contribute to its cytoplasmic localization.
It is paradoxical that YY1 remains in the cytoplasm and yet displays the differential ability to bind DNA through development, with peak activity in very early oocytes and in post-MBT embryos. Because it is unlikely that the extract preparation methods selectively activates DNA binding at these stages, we suggest that maternally derived YY1 is subject to post-translational modifications that occur in the cytoplasm in vivo. It could be that YY1 becomes primed to activate transcription but must receive the appropriate cellular signals to trigger nuclear translocation in specific tissues sometime during later in development.
What is responsible for changes in the DNA binding activity of YY1 during development? There was no apparent correlation between nuclear localization and the ability of YY1 to bind DNA in gel mobility shift assays. DNA binding is probably not regulated by subcellular localization, because YY1 was retained in the cytoplasm throughout the same developmental time frame over which YY1 fluctuates between DNA-binding and non-DNA-binding forms. The ability to reactivate DNA binding activity in late stage oocytes by in vitro phosphatase treatments suggests that YY1 may be negatively regulated in vivo by phosphorylation. However, we found no direct evidence that the phosphorylation state of YY1 changes as oocytes progress through developmental stages when DNA binding activity is lost or regained, and so the significance and mode of fluctuations in DNA binding during early development remain unclear.
In summary, the data presented here suggest that the potential transcriptional activity of YY1 is repressed specifically during oocyte and embryonic development by sequestration to the cytoplasm. The functional significance of changes in the DNA binding activity of YY1 is unclear, and the potential role of YY1 the cytoplasm during development remains to be elucidated. Exclusion of YY1 from nuclei during early development explains the results of expression analyses in which YY1 binding sites had neither stimulatory nor repressive effects on basal transcription from the TK promoter in oocytes and embryos.