Transcriptional selectivity in early mouse embryos: a qualitative study

Genome-wide transcriptional silencing and mRNA stabilization allow the coordinated expression of the meiotic program in mice

The transcriptional dynamic of mammalian cells when these transit from the ubiquitous mitotic to a meiotic-specific program is key to understand this switch central to sexual reproduction. By quantifying active RNA polymerase II and nascent transcripts using single cell dataset and ethynyl-uridine pool-down with sorted cells from synchronized testes, we detailed the transcriptional activity of murine male germ cells. When spermatogonia differentiate, transcription slows down, reaching minimal activity at meiotic entry and resumes during pachytene stage. This event, we termed EMLT (for early meiotic low transcription), is distinct from the silencing of sex chromosomes as it is independent of Setdb1, though it is accompanied by the same chromatin mark, H3K9me3. EMLT is delayed in Stra8KO but occurs in mutants altering meiotic chromosome structure or double-strand break formation or repair. By comparing transcript abundance and nascent transcription we unveil a massive event of messenger RNA stabilization that parallels EMLT. Altogether our data indicate that meiosis is initiated with a nearly silent genome, and we propose that the stabilization of transcripts at that time facilitates the meiotic entry by synchronizing the expression of several meiotic subprograms.

Paracrine in vivo inhibitory effects of hepatitis B virus X protein (HBx) on liver cell proliferation: an alternative mechanism of HBx-related pathogenesis

The role of the hepatitis B virus X protein (HBx) in the pathogenesis of hepatitis B virus (HBV) infection remains unclear. HBx exhibits pleiotropic biological effects, whose in vivo relevance is a matter for debate. In the present report, we have used a combination of HBx-expressing transgenic mice and liver cell transplantation to investigate the in vivo impact of HBx expression on liver cell proliferation and viability in a regenerative context. We show that moderate HBx expression inhibits liver regeneration after partial hepatectomy in HBx-expressing transgenic mice. We also demonstrate that the transplantation of HBx-expressing liver cells, isolated from HBx transgenic mice, is sufficient to inhibit overall recipient liver regeneration after partial hepatectomy. Moreover, the injection of serum samples drawn from HBx-expressing transgenic mice mimicked the inhibitory effect of HBx on liver regeneration. Finally, the incubation of primary rat hepatocytes with the supernatant of HBx-expressing liver cells inhibits cellular DNA synthesis. Taken together, our results demonstrate a paracrine inhibitory effect of HBx on liver cell proliferation and lead us to propose HBV as one of the few viruses implicated in human cancer which act, at least in part, through paracrine biological pathways.

The nucleoskeleton: a permanent structure of cell nuclei regardless of their transcriptional activity

Nuclear matrix or nucleoskeleton is thought to provide structural basis for intranuclear order. However, the nature of this structure is still uncertain because of numerous technical difficulties in its visualization. To reveal the "real" morphology of the nucleoskeleton, and to identify possible sources of structural artifacts, three methods of nucleoskeleton preparations were compared. The nucleoskeleton visualized by all these techniques consists of identical elements: nuclear lamina and an inner network comprising core filaments and the "diffuse" nucleoskeleton. We then tested if the nucleoskeleton is a stable structure or a transient transcription-dependent structure. Incubation with transcription inhibitors (alpha-amanitin, actinomycin D, and DRB) for various periods of time had no obvious effect on the morphology of the nucleoskeleton. A typical nucleoskeleton structure was observed also in a physiological model-in transcriptionally inactive mouse 2-cell embryos and in active 8- to 16-cell embryos. Our data suggest that the nucleoskeleton is a permanent structure of the cell nucleus regardless of the nuclear transcriptional state, and the principal architecture of the nucleoskeleton is identical throughout the interphase.

Somatic linker histone H1 is present throughout mouse embryogenesis and is not replaced by variant H1 degrees

A striking feature of early embryogenesis in a number of organisms is the use of embryonic linker histones or high mobility group proteins in place of somatic histone H1. The transition in chromatin composition towards somatic H1 appears to be correlated with a major increase in transcription at the activation of the zygotic genome. Previous studies have supported the idea that the mouse embryo essentially follows this pattern, with the significant difference that the substitute linker histone might be the differentiation variant H1 degrees, rather than an embryonic variant. We show that histone H1 degrees is not a major linker histone during early mouse development. Instead, somatic H1 was present throughout this period. Though present in mature oocytes, somatic H1 was not found on maternal metaphase II chromatin. Upon formation of pronuclear envelopes, somatic H1 was rapidly incorporated onto maternal and paternal chromatin, and the amount of somatic H1 steadily increased on embryonic chromatin through to the 8-cell stage. Microinjection of somatic H1 into oocytes, and nuclear transfer experiments, demonstrated that factors in the oocyte cytoplasm and the nuclear envelope, played central roles in regulating the loading of H1 onto chromatin. Exchange of H1 from transferred nuclei onto maternal chromatin required breakdown of the nuclear envelope and the extent of exchange was inversely correlated with the developmental advancement of the donor nucleus.

Mouse embryos do not wait for the MBT: chromatin and RNA polymerase remodeling in genome activation at the onset of development

In Xenopus and Drosophila embryos, activation of the zygotic genome occurs after a series of rapid nuclear divisions in which DNA replication occupies most of the cell cycle. In these organisms, it has been proposed that zygotic transcription does not begin until a threshold nucleocytoplasmic ratio has been obtained in which repressive factors are titrated out and interphase becomes long enough to allow synthesis of transcripts. In mammalian embryos, however, a model of threshold nucleocytoplasmic ratios does not seem to apply, as beginning with the 1-cell stage, there are regulated cell cycles with the expression of zygotic transcripts during the cleavage period. By taking advantage of the slower kinetics at the onset of mouse development, we have characterized changes in chromatin structure and the basal transcription machinery throughout the transition from transcriptional incompetence, to minor activation of the zygotic genome during the 1-cell stage, and through major genome activation at the 2-cell stage. Further maturation of chromatin structure continues through subsequent cleavage cycles as a foundation for the first cellular differentiations in the blastocyst. The epigenetic chromatin modifications that occur during the cleavage period may have long range and inheritable effects and are undoubtedly important in the ability of the mammalian oocyte to remodel previously defined nuclear structures and cell fates.

Efficient selection of transgenic mouse embryos using EGFP as a marker gene

We have established a reliable method that uses the EGFP (Enhanced Green Fluorescent Protein) gene as a marker for selecting transgenic embryos from preimplantation embryos. Embryos that were subjected to the pronuclear microinjection of the CMV/beta-actin/EGFP fusion gene were cultured in vitro until they developed into the morulae- or blastocyst-stage. The expression of EGFP was easily observed by a fluorescent microscopy. There appeared to be no damage to the in vivo developmental ability of the embryos in response to the EGFP excitation light, which utilized an IB filter for a period of 30 min. Modified PCR analysis using Dpn I and Bal 31 digestion of the embryonic DNA showed that all of the embryos expressing EGFP in all their cells were transgenic, while more than half with mosaic expression of EGFP were not transgenic. Approximately 77% of pups born from the embryos that uniformly expressed the EGFP gene were transgenic, while 21.4% of pups from the embryos with mosaic expression were transgenics. The results showed that the use of EGFP as a marker is very useful and reliable for selecting transgenic embryos, and that it is important to transfer the embryos expressing EGFP in all their cells to obtain truly transgenic animals.

Relief of a repressed gene expression state in the mouse 1-cell embryo requires DNA replication

In the mouse, transcriptional permissiveness is established in the fertilized egg prior to the activation of zygotic genes at the 2-cell stage. Therefore, gene inactivity initiated at the end of gametogenesis results from a complex process, involving more than an inhibition of the basal transcriptional apparatus. We have examined the ability of the first intron (I1) of the human hypoxanthine phosphoribosyl transferase gene, which functions as an enhancer in embryonic stem cells, to activate a reporter gene when placed proximally to or at a distance from the HSV-tk promoter, or when integrated into the mouse genome as part of a stable transgene. In microinjected embryos, I1 functions as an enhancer sequence; however, its competence for long-range activation appears only after the late 1-cell stage and depends on the first DNA replication. Moreover, activation of microinjected transgenes from proximal enhancers occurs in the late 2-cell embryo and in the male pronucleus of 1-cell embryos blocked for DNA replication; whereas, for integrated transgenes, proximal enhancer activity is subject to position effects in the 2-cell embryo and first occurs at the 2- or 4-cell stage, but only after completion of DNA replication. Therefore, the absence of long-range activation and a non-permissive genomic state (the relief of which both depend on DNA replication), together with an inactive transcriptional apparatus, appear to converge to prevent any gene activity in the 1-cell embryo. We propose that the embryo exploits the process of DNA replication to relieve the transcriptionally repressive state that was initially established to fulfil two purposes: (1) to arrest maternal gene expression in the maturing oocyte and (2) to protect the unicellular egg and 1-cell embryo from premature differentiation. Reactivation of gene expression by DNA replication would therefore serve to coordinate cell proliferation and differentiation in the preimplantation embryo.

Quantitative analysis of luciferase activity of viral and hybrid promoters in bovine preimplantation embryos

This study was conducted to investigate quantitatively the luciferase activity of gene constructs with viral and hybrid enhancers and promoters in bovine preimplantation embryos by using firefly luciferase reporter genes. In Experiment I, to examine the stability of the luciferase, bioluminescence intensity of bovine embryos injected with the luciferase gene driven by the SV40 early promoter and enhancer (SVEluc) was measured with a luminometer at 2 days after microinjection. The results indicated that the bioluminescence could be analysed at any time within 30 min because the luciferase activity was constant during the measurement period from 5 to 30 min. In Experiment II, the luciferase expression of fertilized oocytes injected with four gene constructs (TKEluc, TK6WEluc, SVEluc, and Miwluc) was analysed by using a photon imaging system at 2 or 6 days following microinjection. The results from Experiment II indicated that the reporter gene governed by the Miw promoter (RSV LTR and chicken beta-actin promoter) was expressed more intensively in bovine morulae and blastocysts than three other gene constructs. In Experiment III, the effect of SV40 enhancer was investigated when fused downstream to the luciferase cDNA of the Miwluc vector. The results showed that SV40 enhancer further activated the luciferase activity of the Miw promoter in bovine preimplantation embryos. It was concluded, therefore, that the Miw promoter together with the SV40 enhancer would confer the strongest expression of the firefly luciferase reporter gene among the gene constructs tested in preimplantation bovine embryos.

Developmental activation of an episomic hsp70 gene promoter in two-cell mouse embryos by transcription factor Sp1

To investigate the control of zygotic genome expression in two-cell mouse embryos, we studied transcription factors required for transient expression of microinjected DNA constructs driven by the promoter of one of the earliest genes activated after fertilization in this system, the heat shock gene hsp70. Cis-acting elements required for hsp70 activation were first investigated by mutational analysis. Mutation of the TATA box and a proximal GC box strongly inhibited construct expression, while that of a CCAAT box had no effect. Transcription factors binding the wild-type hsp70 promoter were then titrated in vivo by coinjecting the construct with double-stranded oligodeoxyribonucleotides containing definite consensus sequences. Wild-type GC box oligonucleotides strongly inhibited construct expression, while those containing mutated GC boxes, wild-type CCAAT boxes, and heat shock elements had no effects. Finally, construct expression was challenged by coinjecting antibodies to specific transcription factors. Antibodies to factor Sp1 depressed construct expression in a dose-dependent manner, while those to Sp2, HSF1 and HSF2 were ineffective. These results pinpoint the Sp1 transcription factor as an absolute requirement for activation of the hsp70 gene promoter in two-cell mouse embryos, and make this factor a candidate for a major regulator of the onset of murine zygotic genome expression.

cis-acting elements that mediate the negative regulation of Moloney murine leukemia virus in mouse early embryos

We have addressed the question of the nature of Moloney murine leukemia virus (MoMuLV) repression in mouse embryos by assaying for the transient expression of MoMuLV-derived constructs microinjected into early cleavage embryos. We show that the same cis-acting DNA sequences responsible for the block in MoMuLV expression in embryonal carcinoma cell lines operate in early embryos: (i) the MoMuLV long terminal repeat is nonfunctional, and (ii) the +147 to +163 repressor binding site, or negative regulatory element, negatively regulates the expression from an active promoter.

Progressive maturation of chromatin structure regulates HSP70.1 gene expression in the preimplantation mouse embryo

In the widely studied model organisms, Drosophila and Xenopus, early embryogenesis involves an extended series of nuclear divisions prior to activation of the zygotic genome. The mammalian embryo differs in that the early cleavage phase is already characterized by regulated cell cycles with specific zygotic gene expression. In the mouse, where major activation of the zygotic genome occurs at the 2-cell stage, the HSP70.1 gene is among the earliest genes to be expressed. We investigated the developmentally regulated expression of this gene during the preimplantation period, using a luciferase transgene, with or without flanking scaffold attachment regions (SARs). Cleavage stage-specific modifications in expression profiles were examined in terms of histone H4 acetylation status, topoisomerase II activity, and the localisation of HMG-I/Y, a nuclear protein with known affinity for the AT-tracts of SARs. We demonstrate that HSP70.1-associated transcription factors are not limiting, and that instead, there is a progressive maturation of chromatin structure that is directly involved in HSP70.1 regulation during early mouse development.

A unique role for enhancers is revealed during early mouse development

Transcription and replication of genes in mammalian cells always requires a promoter or replication origin, respectively, but the ability of enhancers to stimulate these regulatory elements and the interactions that mediate this stimulation are developmentally acquired. The primary function of enhancers is to prevent repression, which appears to result from particular components of chromatin structure. Factors responsible for this repression are present in the maternal nucleus of oocytes and its descendant, the maternal pronucleus of mouse 1-cell embryos and in mouse 2-cell embryos, but are absent in the paternal pronucleus. Thus, enhancers are not needed to achieve efficient transcription and replication in paternal pronuclei. However, enhancers, even in the presence of their specific activation protein, are inactive prior to formation of a 2-cell embryo, suggesting that a coactivator essential for enhancer function is not available until zygotic gene expression begins. Furthermore, enhancer stimulation of transcription appears to be mediated through a promoter transcription factor, but this interaction can change as cells undergo differentiation, switching from a TATA-box independent to a TATA-box dependent mode.

A TATA-less promoter containing binding sites for ubiquitous transcription factors mediates cell type-specific regulation of the gene for transcription enhancer factor-1 (TEF-1)

TEF-1 is a tissue-specific human transcription factor which binds to and activates transcription from the SV40 early promoter and the HPV-16 E6/E7 promoter and may be involved in regulation of muscle-specific and placenta-specific gene expression. To investigate the mechanism of its tissue-specific expression, we have isolated up to 3 kilobase pairs of 5'-flanking DNA and characterized the promoter of the gene for TEF-1. Multiple transcription start sites centering on a motif similar to the initiator element (Inr) were identified. A minimal promoter, which contains no recognizable TATA element but contains an Inr, delimited at -137 base pairs had full transcriptional activity both in vivo in HeLa cells and in vitro in HeLa cell extracts. This promoter is also highly active in vitro in lymphoid cell extracts, but not in vivo in lymphoid cell lines, which do not express the endogenous TEF-1 gene. The minimal promoter, which is sufficient to direct tissue-specific expression of the TEF-1 gene in vivo, contains multiple sites which bind the ubiquitous transcription factors Sp1 and ATF-1. Mutation of the Inr completely abolished transcription from the major start site while transcription from the minor sites was slightly augmented. Inactivation of the proximal Sp1 site abolished transcription from the principle start site and increased transcription from a 5' minor start site. Insertion of a TATA box element did not qualitatively alter the pattern of start site usage which seemed to be dependent upon integrity of the upstream Sp1 site. These observations suggest a "cross-talk" between the Inr and a proximal element to fix transcription start sites, which is independent of spacing and the presence of a TATA element.

Expression of the HSP 70.1 gene, a landmark of early zygotic activity in the mouse embryo, is restricted to the first burst of transcription

Activation of the mouse embryonic genome at the 2-cell stage is characterized by the synthesis of several alpha-amanitin-sensitive polypeptides, some of which belong to the multigenic hsp 70 family. In the present work we show that a member of this family, the HSP 70.1 gene, is highly transcribed at the onset of zygotic genome activation. Transcription of this gene began as early as the 1-cell stage. Expression of the gene continued through the early 2-cell stage but was repressed before the completion of the second round of DNA replication. During this period we observed that the level of transcription was modulated by in vitro culture conditions. The coincidence of repression of HSP70.1 transcription with the second round of DNA replication was not found for other transcription-dependent polypeptides synthesized at the 2-cell stage.

Onset of paternal gene activation in early mouse embryos fertilized with transgenic mouse sperm

We investigated the onset of paternal gene expression in the early mouse embryo. We obtained transgenic mouse embryos by fertilizing BD (C57BL/6N x DBA) F1 hybrid female oocytes in vitro, with sperm from homozygous transgenic males carrying integrated chicken beta-actin promoter-driven firefly luciferase cDNA. We then examined the RNA and protein synthesis of the luciferase gene in embryos from the 1- to 2-cell stage. RNA transcripts of the luciferase gene were first detected in the 1-cell stage embryos as early as 13 hr postinsemination, just prior to elongation. By photon-count imaging, functional luciferase was identified at the 2-cell stage 23 hr postinsemination. These findings indicate that the paternal endogenous gene is already transcribed in the late 1-cell embryos, although paternally derived protein is not synthesized until the 2-cell stage. Therefore, these results suggest that the embryonic gene is activated as early as the late 1-cell stage.

Regulation of gene expression in the mouse oocyte and early preimplantation embryo: developmental changes in Sp1 and TATA box-binding protein, TBP

We previously demonstrated that an Sp1-dependent reporter gene is preferentially expressed in G2 of the 1-cell mouse embryo following microinjection of the male pronucleus when compared to microinjection of the female pronucleus (P.T. Ram and R.M. Schultz, 1993, Dev. Biol. 156, 552–556). We also noted that expression of the reporter gene is not observed following microinjection of the germinal vesicle of the fully grown oocyte. In the present study, we examined expression of this reporter gene during oocyte growth, as well as the nuclear concentration of two transcription factors, Sp1 and the TATA box-binding protein, TBP, during oocyte growth and the first cell cycle. The extent of reporter gene expression decreases during oocyte growth and this decrease correlates with the decrease in nuclear concentration of Sp1, as determined by confocal immunofluorescent microscopy. In addition, results of immunoblotting experiments also indicate a similar decrease in the total concentration of Sp1 during oocyte growth. The nuclear concentration of TBP also decreases during oocyte growth, as determined by confocal immunofluorescent microscopy. Following fertilization, the pronuclear concentration of these two transcription factors increases in a time-dependent fashion and the concentration of each is greater in the male pronucleus as compared to the female pronucleus. For each pronucleus and for each transcription factor, this increase in nuclear concentration is inhibited by aphidicolin, which inhibits DNA synthesis. Last, the increase in nuclear concentration of these two proteins observed between the 1-cell and 2-cell stages does not require transcription or cytokinesis.

Regulation of zygotic gene activation in the mouse

Zygotic gene activation (ZGA) is the critical event that governs the transition from maternal to embryonic control of development. In the mouse, ZGA occurs during the 2-cell stage and appears to be regulated by the time following fertilization, i.e. a zygotic clock, rather than by progression through the first cell cycle. The onset of ZGA must depend on maternally inherited proteins, and post-translational modification of these maternally derived proteins is likely to play a role in ZGA. Consistent with this prediction is that protein phosphorylation catalyzed by the cAMP-dependent protein kinase is involved in ZGA and that protein synthesis is not required for ZGA. Recent results suggest that ZGA may occur earlier than previously thought, i.e. not during the 2-cell stage, but rather in G2 of the 1-cell embryo. Thus ZGA may comprise a period of minor gene activation in the 1-cell embryo that is followed by a period of major gene activation in the 2-cell embryo. Following ZGA, the expression of constitutively activated genes may require an enhancer.

Transient expression of a human beta-actin promoter/lacZ gene introduced into mouse embryos correlates with a low degree of methylation

We have analysed the correlation between expression and methylation for the human beta-actin promoter introduced into mouse embryos. The beta-actin promoter was fused to the reporter gene lacZ, and expression was analysed after pronuclear injection into fertilized mouse eggs. We analysed transient expression in in vitro cultured preimplantation embryos and expression after chromosomal integration in 5 independent lines of transgenic mice. The in vitro cultured preimplantation embryos expressed lacZ from the 2-cell to the blastocyst stages, and most abundantly at the morula stage. By increasing the amount of injected DNA, a larger proportion of embryos expressed lacZ. Embryos expressing lacZ in only a subset of the blastomeres were detected at all preimplantation stages. In contrast to the transient expression after injection, we have not detected lacZ expression in any of the 5 analysed lines of transgenic mice carrying the same construct. The lack of expression in transgenic mice correlates with hypermethylation of C residues in the vast majority of CG sequences in the integrated beta-actin/lacZ construct, whereas the injected construct was completely nonmethylated. We discuss methylation and other possible reasons for the observed differences in expression between injected and integrated copies of the beta-actin/lacZ construct and for lacZ expression in only a subset of blastomeres in preimplantation embryos.

Expression of microinjected DNA and RNA in early rabbit embryos: changes in permissiveness for expression and transcriptional selectivity

Gene expression in rabbit early development was investigated by microinjecting LacZ DNA and LacZ RNA in 1-cell and 2-cell embryos. Expression of LacZ DNA could not be obtained before 30-36 hpf, although synthetic LacZ RNA was translated from 12 hpf at the least. The onset of expression of microinjected DNA correlated with the 8- to 16-cell stage. This suggests that before this stage, there is a general negative control of gene expression. The arrest of in vitro development at the 2- to 8-cell stages did not inhibit LacZ expression, which still occurred at 33 hpf. In addition the inhibition of the first cleavage by nocodazole resulted in LacZ expression in 1-cell embryos. Expression of microinjected DNA thus occurs at a fixed time after fertilization and is independent of cleavages and of the second and subsequent DNA replications. Therefore, the changes in permissiveness for the expression of microinjected DNA in rabbit embryos are reminiscent of those in mouse embryos. Transcriptional selectivity in rabbit embryos was compared to that in early mouse embryos. In both species, Sp1-sensitive promoters were active and the promoter of simian virus 40 did not require far upstream enhancers before late cleavage stages; genes driven by the -447, +563 region of murine leukemia virus were repressed. In rabbit, however, the H-2Kb promoter active in mouse was silent. Altogether, the results illustrate a remarkable conservation of the characteristics of the transcription in early rabbit and mouse embryos and the independence of its resumption from the pattern of cleavage.

Capture of a cellular transcriptional unit by a retrovirus: mode of provirus activation in embryonal carcinoma cells

The expression of murine leukemia provirus in embryonal carcinoma (EC) cells is blocked by a mechanism still incompletely understood. The blockage is not overcome by deleting a large portion of the enhancer region (in U3) in recombinant retroviruses (M-MuLVneo delta Enh). This confirms the presence of negative elements outside the viral 82-bp repeats. However, a few sites in the genomes of EC cells permit M-MuLVneo delta Enh proviral expression. One such site, identified in PCC4, PCC3, and LT, was studied. The complete analysis of the mechanism of activation by Northern (RNA) blotting, cloning, and sequencing of partial cDNA copies of the viral transcript and of the site of integration establishes that viral transcripts are initiated from an upstream host-cell promoter and are spliced from a host donor to a cryptic viral acceptor at position 542 in the Moloney murine leukemia virus (M-MuLV) genome. In consequence, the mature transcripts are host cell-virus fusion transcripts from which M-MuLV sequences, including the cis-active negative elements of the 5' long terminal repeat-containing region, are absent. The provirus integrates apparently randomly into any of the three most proximal introns of the transcriptional unit. The host cell promoter contains a TATA box and 14 potential SpI binding sites included in a 1.0-kb GC-rich island. These elements promote gene expression of recombinant vectors in EC and differentiated cells. The mechanism described points to a mechanism by which retroviruses can be transcribed from upstream nonviral elements and can acquire host genes by 5' annexation of exons.