Gene expression in mouse oocytes and preimplantation embryos: use of suppression subtractive hybridization to identify oocyte- and embryo-specific genes

The paucity of biological material has inhibited identifying genes that are differentially expressed during mammalian oogenesis and preimplantation development. We report here the linear amplification of mRNA from small numbers of mouse oocytes and preimplantation embryos to generate amounts of sense RNA that are sufficient for suppression subtractive hybridization. The resulting oocyte-specific and 8-cell-specific cDNA libraries were partially characterized, and the known oocyte-specific ZP1, ZP2, GDF-9, BMP15, and H1(oo) genes were found in the oocyte-specific cDNA library but not in the 8-cell-specific library. Further characterization of the subtracted oocyte and 8-cell embryo cDNA libraries should furnish a trove of information regarding temporal changes in gene expression during oogenesis and preimplantation development in the mouse.     

Differential screening of a cDNA library with cDNA probes amplified in a heterologous host: isolation of murine GRP78 (BiP) and other serum-regulated low-abundance mRNAs

A novel method was used to screen differentially a cDNA library for clones representing serum-regulated mRNA species of low abundance. To increase the amount of probe available for screening, the cDNA probe was cloned and amplified. Two separate cDNA 'probe' libraries were constructed in the Escherichia coli plasmid vector pDE613, using poly(A)+mRNA from murine cells at 0 and 16 h after stimulation of a G0 population. Radiolabelled plasmid DNA from each library was hybridized sequentially to colony blots of the third 'target' library, constructed with mRNA from serum-stimulated cells in the Bacillus subtilis vector pBD214. Differential screening of the target cDNA library with the two probe libraries identified novel murine cDNA clones, some representing cytoplasmic poly(A)+RNA species of low (0.01%) abundance, accumulating after serum stimulation of a quiescent mouse embryo fibroblast population. One cDNA clone was found to correspond to mitochondrial 16S rRNA and a second was identified as the murine equivalent of previously described cDNA clones for the hamster 78-kDa glucose-regulated protein (GRP78) and the rat immunoglobulin heavy-chain-binding protein. GRP78 mRNA has not previously been recognized as a serum-inducible message.     

Gene expression in chemically transformed mouse embryo cells: selective enhancement of the expression of C type RNA tumor virus genes.

Treatment of a nontumorigenic clone of AKR mouse embryo cells in culture with a variety of polycyclic aromatic hydrocarbons has resulted in the development of derivative clones which are highly tumorigenic and exhibit other characteristics of the transformed phenotype. A 3-methylcholanthrene-transformed derivative clone (clone MCA) has been compared to the parent clone (clone 2B) with respect to the abundance and diversity of polysomal poly(A)-containing mRNA sequences. Hybridization kinetic experiments show that the poly(A)-containing sequences of both clones are organized into indistinguishable abundance classes, and that the vast majority of the sequences are common to both the parent and derivative clones. The levels of two specific messenger RNAs (α- and β-globin mRNA) which characterize highly differentiated mouse erythroid cells were much less than 1 molecule per cell in either cell type. Titration of a balanced complementary DNA probe to AKR murine leukemia virus (AKR-MuLV) 70S RNA with purified polysomal poly(A)-containing RNA from both parent and derivative clones shows that approximately 5000 and 1200 viral 35S RNA equivalents are present in the cytoplasm of growing and resting clone MCA cells, respectively. Rapidly growing clone 2B cells contain less than about 30 viral 35S RNA equivalents per cell. Viral specific sequences therefore correspond to members of the high abundance class of poly(A)-containing RNA sequences in clone MCA cells and to the low abundance class of sequences in clone 2B cells. Within the limits of detection, this large increase in abundance is characteristic only of viral specific RNA sequences.     

Isolation of endo A cDNA from mouse 8-cell stage embryos

To analyse the species of genes expressed in a cleavage stage mouse embryo, we have constructed a cDNA library containing 3.0 x 10(5) independent clones from about 2 x 10(3) embryos at the 8-cell stage of development. Endo A cDNA prepared from parietal yolk sac endoderm like PYS-2 cells was used to screen the library. Southern blot analyses using the endo A sequence as a probe and restriction mapping analyses revealed that four independent recombinants had been inserted endo A sequence. Sequencing data of these clones showed that endo A mRNA present in the 8-cell stage embryo is identical to that of parietal yolk sac endoderm cells.     

Quantification of basigin mRNA in mouse oocytes and preimplantation embryos by competitive RT-PCR

Basigin is a member of the immunoglobulin superfamily and a key molecule related to mouse blastocyst implantation. Whether preimplantation mouse embryos express basigin mRNA is still unknown. The aim of this study was to use a quantitative competitive polymerase chain reaction to assess quantitatively the levels of basigin mRNA in mouse oocyte and preimplantation embryos. Basigin mRNA was detected in the oocyte and all the stages of preimplantation embryos. The levels of basigin mRNA were 0.0606 +/- 0.0282 in the oocyte, 0.0102 +/- 0.0036 in the zygote, 0.0007 +/- 0.0003 in the 2-cell embryo, 0.0031 +/- 0.0017 in the 4-cell embryo, 0.0084 +/- 0.0024 in the 8-cell embryo, 0.0537 +/- 0.0121 in the morula and 0.0392 +/- 0.0161 attomoles in the blastocyst, respectively. The levels of basigin mRNA in the oocyte, morula and blastocyst were significantly higher than those in the zygote and embryos at the 2-cell, 4-cell and 8-cell stages. The high level of basigin expression in the blastocyst may play a role during embryo implantation.     

Disruption of the cytokeratin filament network in the preimplantation mouse embryo

The distribution of the cytokeratin network in the intact preimplantation mouse embryo and the role of cytokeratin filaments in trophectoderm differentiation were investigated by means of whole-mount indirect immunofluorescence microscopy and microinjection of anti-cytokeratin antibody. Assembled cytokeratin filaments were detected in some blastomeres as early as the compacted 8-cell stage. The incidence and organization of cytokeratin filaments increased during the morula stage, although individual blastomeres varied in their content of assembled filaments. At the blastocyst stage, each trophectoderm cell contained an intricate network of cytokeratin filaments, and examination of sectioned blastocysts confirmed that extensive arrays of cytokeratin filaments were restricted to cells of the trophectoderm. Microinjection of anticytokeratin antibody into individual mural trophectoderm cells of expanded blastocysts resulted in a dramatic rearrangement of the cytokeratin network in these cells. Moreover, antibody injection into 2-cell embryos inhibited assembly of the cytokeratin network during the next two days of development. Despite this disruption of cytokeratin assembly, the injected embryos compacted and developed into blastocysts with normal morphology and nuclear numbers. These results suggest that formation of an elaborate cytokeratin network in preimplantation mouse embryos is unnecessary for the initial stages of trophectoderm differentiation resulting in blastocyst formation.     

DDX21 is a p38-MAPK-sensitive nucleolar protein necessary for mouse preimplantation embryo development and cell-fate specification

Successful navigation of the mouse preimplantation stages of development, during which three distinct blastocyst lineages are derived, represents a prerequisite for continued development. We previously identified a role for p38-mitogen-activated kinases (p38-MAPK) regulating blastocyst inner cell mass (ICM) cell fate, specifically primitive endoderm (PrE) differentiation, that is intimately linked to rRNA precursor processing, polysome formation and protein translation regulation. Here, we develop this work by assaying the role of DEAD-box RNA helicase 21 (DDX21), a known regulator of rRNA processing, in the context of p38-MAPK regulation of preimplantation mouse embryo development. We show nuclear DDX21 protein is robustly expressed from the 16-cell stage, becoming exclusively nucleolar during blastocyst maturation, a localization dependent on active p38-MAPK. siRNA-mediated clonal Ddx21 knockdown within developing embryos is associated with profound cell-autonomous and non-autonomous proliferation defects and reduced blastocyst volume, by the equivalent peri-implantation blastocyst stage. Moreover, ICM residing Ddx21 knockdown clones express the EPI marker NANOG but rarely express the PrE differentiation marker GATA4. These data contribute further significance to the emerging importance of lineage-specific translation regulation, as identified for p38-MAPK, during mouse preimplantation development.     

Variations in the amounts of RNA polymerase forms I, II and III during preimplantation development in the mouse.

DNA-dependent RNA polymerase has been measured at various stages of preimplantation development in mouse embryos. The total RNA polymerase activity per embryo increases rapidly from the 8-cell stage to the blastocyst stage. Studies with low α-amanitin concentrations, which inhibit form II RNA polymerase, and high α-amanitin concentrations, which inhibit both form II and III RNA polymerases indicate that the relative proportions of the three forms change significantly during preimplantation development. The changes which occur in the types and levels of RNA polymerase appear to parallel corresponding changes in the synthesis of the major classes of RNA.     

Differential splicing of thymosin beta 4 mRNA

A cDNA clone was isolated from a mouse pre-B cell line, the sequence of which has a very high homology with rat and human thymosin beta 4 genes. However, the mouse clone has an insertion of 98 bp relative to the published rat and human sequences upstream of the coding region. By isolation of a second set of clones from a different cDNA library and by cloning a PCR amplified region of mouse genomic DNA it was confirmed that the insertion is not a cloning artifact. Furthermore, it was shown by RNase protection assays with RNA from the pre-B cell line that two sizes of thymosin beta 4 mRNA exist, a long form containing the 98 nucleotide insertion, and a short form that corresponds to the known rat and human mRNA. The short form is about 50 times more abundant than the long form. Analysis of genomic DNA by sequencing and Southern blotting revealed that both forms are encoded by a single gene in the mouse. The two forms of mRNA arise by differential RNA splicing; the long mRNA contains three separate exons, whereas the short mRNA is missing exon 2. The long mRNA is present in two different pre-B cell lines, spleen and thymus, but could not be detected in brain, liver, and kidney. It is possible that the longer mRNA, which encodes a hydrophobic NH2-extension of six additional amino acids, plays a role in lymphocyte function or development. In contrast to the mouse which has a single thymosin beta 4 gene, rat and human have multiple homologs. Most or all of these also contain sequences that cross-hybridize with the newly discovered exon 2. A polymorphic thymosin beta 4 gene has been found in human DNA.