| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Embryo |
a Laboratoire de Biologie du Développement et Biotechnologie, INRA, 78352 Jouy en Josas Cedex, France
b Laboratoire de Mathématiques, Informatique et Génome, INRA, Route de Saint-Cyr, 78026 Versailles Cedex, France
c Laboratoire de Radiobiologie et d'Etude du Génome, INRA, 78352 Jouy en Josas Cedex, France
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
developmental biology, early development, embryo, gene regulation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Structural analysis of chromatin modifications has progressed significantly over the past few years; nevertheless, identification of the first zygotic transcripts synthesized by the newly transcriptionally competent genome is still required in order to understand how the silent genome becomes active at the beginning of embryonic life. Several major questions remain: Are all the genes suddenly turned on, and then progressively and specifically repressed to establish a cell lineage-specific pattern of gene expression, or is a subset of genes specifically transcribed at that stage of development? Are the first embryonic transcripts absolutely necessary for subsequent steps of development due to the appearance of new functions in the cells or due to the sudden lack of maternal homologous molecules? These still-unanswered questions require analysis of the embryonic transcriptome just after ZGA. Mouse embryos develop to the 2-cell stage in the absence of embryonic transcription, whereas in other mammals, activation of zygotic genes is more progressive and the blockade occurs after several cell cycles (at the 8-cell stage in rabbits and cows, and at the 4-cell stage in humans) [10–13]. We have thus concluded that rabbits constitute a good experimental model for which to examine the onset of zygotic transcription.
Before polymerase chain reaction (PCR)-based technologies were improved, transcriptome analyses in mammals were impaired by the very small amounts of available material (i.e., a few picograms of messenger RNA per embryo). In a pioneering work, Taylor and Piko [14] showed that embryonic transcription in mammals results not only in a renewal of molecules already present as maternal factors in the embryo, but that it also produces a high proportion of new molecular species. Rothstein et al. [15] later established a subtracted library potentially enriched in the mouse first zygotic transcripts. Through differential display reverse transcription (RT)-PCR, a few studies were more informative, but they concerned only a few transcripts [16–19]. Recently, Ma et al. [20] used mRNA differential display to conduct a more global analysis (200 amplicons) of gene expression at the 2-cell stage in mice. Further studies are needed to identify the corresponding transcripts. A recent and large-scale cDNA analysis of expressed sequence tags (ESTs) from mouse embryos was carried out to obtain expression patterns from thousands of genes during preimplantation development [21]. Such a method relies entirely on the frequency with which individual sequences are found in cDNA libraries obtained by PCR at each stage. Such a global analysis provides many data from mice and humans because of the large number of sequences already available in the databases.
Despite a growing interest in using the rabbit as a model for human genetics [22], only a few cDNA sequences are available for this species in the databases, which precludes the establishment of a microarray without previous isolation of cDNAs. This situation is encountered in many other species, but alternative methods such as suppression subtractive hybridization (SSH) [23] allow the identification of genes involved in specific biological situations. In this study, we used this technique because it was compatible with scarce material in order to study gene expression at ZGA in the rabbit.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Superovulation was induced in New Zealand White rabbits by five s.c. injections of FSH/LH (Stimufol, Mérial, France) for 3 days before mating (5 µg FSH/1 µg LH on Day 1, 5 µg FSH/1 µg LH and 10 µg FSH/2 µg LH on Day 2, and 10 µg FSH/2 µg LH and 5 µg FSH/1 µg LH on Day 3). Rabbits were mated 12 h after the last FSH/LH injection (time = 0 postcoitum; pc), and 30 IU of hCG (Chorulon, Intervet, Beaucouzé, France) was injected a few minutes after mating. At 19 (or 24) h pc the rabbits were killed, and the oviducts were recovered and maintained at 37°C until perfusion with PBS to collect the embryos. One-cell-stage embryos were recovered 19 h pc and immediately frozen. Other embryonic stages were obtained by in vitro culture in Medium 199 (Gibco BRL, Cergy Pontoise, France) supplemented with 10% fetal calf serum in an atmosphere of 5% CO2 at 38°C up to the desired stages. Four-cell embryos were obtained by an 8-h in vitro culture of embryos recovered at the 2-cell stage at 24 h pc. Embryos at the 8- to 16-cell, morula, and blastocyst stages were obtained from 19 h pc 1-cell-stage embryos cultured in vitro for 31, 50, and 78 h, respectively. In order to determine whether RNAs were zygotic or maternal transcripts, some embryos were cultured from the 1-cell stage onward (19 h pc) in a medium containing 100 µg/ml of 
-amanitin and were collected at the same time as untreated 8- to 16-cell embryos.
RNA Purification, Reverse Transcription,> cDNA Amplification, and Subtraction
Typically, SSH requires 2 µg of poly(A)+ RNA from each stage [23]. Because rabbit embryos at these stages contain only 
50–100 pg of poly(A)+ RNA, we preamplified the total cDNA population of both stages by RT-PCR using the SMART PCR cDNA Synthesis kit (Clontech Laboratories, Palo Alto, CA), which allows amplification of cDNA templates by long-distance PCR. Total RNA from 500 1-cell or morula stage embryos were extracted using the guanidium isothiocyanate method with the RNA-PLUS kit (Q. Biogen, Illkirch, France) as described by the manufacturer. Five micrograms of glycogen (Boehringer-Mannheim, Mannheim, Germany) were used as a carrier. RNAs were resuspended in 3.5 µl of RNase-free water. All purified total RNAs (approximately 250 ng for each stage) were immediately used for cDNA synthesis using the SMART cDNA synthesis kit (Clontech). The final volume for reverse transcription was 10 µl; 4 µl was stored at -20°C (to generate probes for differential screening, see below), and the remainder was diluted in 24 µl of TE buffer (10 mM Tris pH 7.6, 1 mM EDTA). Six microliters of diluted cDNA were subjected to 20 PCR cycles. SSH was performed with the Clontech PCR-Select cDNA subtraction kit as described by the manufacturer.
Cloning in the TA Vector and Generating> the Replica Membranes
The subtracted library was cloned into pCR 2.1 as described by the manufacturer (Original TA cloning kit; Invitrogen, Cergy Pontoise, France). Approximately 30 ng of PCR-amplified cDNA was ligated without further purification into 50 ng of vector at 14°C overnight, and the ligation mixture was transformed into TOP10F' (Invitrogen) competent cells. The transformants were spread on Biodyne A nylon membranes (Pall, Portsmouth, England) laying on LB agar plates containing 50 µg/ml of ampicillin. Bacteria were grown until colonies were visible at a density of 250 colonies per dish (150 mm in diameter). Two replica membranes were generated from each plate on Biodyne A membranes.
Differential Screening of the Subtracted Library
Two probes corresponding to 1-cell and morula stages were constructed using the nondiluted reaction mixture of the first-strand cDNAs generated for subtraction. These cDNAs were amplified through SMART global PCR as described by the manufacturer for cDNA library construction, meaning they were not digested with RsaI. About 50 ng of amplified cDNA from each embryonic stage were labeled by random priming (3–4 x 108 cpm/µg). The two replica membranes of six Petri dishes were prehybridized for 24 h in Church buffer [24] at 65°C and hybridized for 48 h with these probes diluted at 2 x 106 cpm/ml in Church buffer. Membranes were washed twice in 2x standard sodium citrate (SSC)/0.1% SDS at 65°C for 15 min and twice in 0.2x SSC/0.1% SDS for 15 min. Membranes were autoradiographed at -80°C for 72 h with an amplifier screen. A second exposure was realized for 3 wk in order to better detect 1-cell-stage labeling. Results obtained with both probes were compared in order to classify the colonies according to their labeling.
DNA Sequencing and Sequence Analysis
The 114 inserts pointed out by differential screening were sequenced. We used the standard nucleotide-nucleotide Basic Local Alignment Search Tool (blastn) program maintained by the National Center for Biotechnology Information (NCBI) to compare sequences of isolated clones with the nonredundant EST database (dbEST) and the high throughput genomic sequences (HTGS) nucleic acid databases. We used the NCBI translated database (tblastx) to make amino acid comparisons and the PROSITE database of protein families and domains for motif and domain comparisons.
Embryo and Tissue Slot Blot Analysis
Total RNA from 50 embryos at each stage was extracted by the same method described for library construction, except that 10 µg of 5S RNA were added with RNA-PLUS in order to define the recovery rate by measuring optic density at 260 nm. Tissue RNAs were purified by the LiCl/urea method. Global RT-PCR was carried out as described in [25] with a few modifications. Reverse transcription was oligo(dT) primed and carried out using a mix of Superscript II (100 units) and avian myeloblastosis virus (1 unit) reverse transcriptases (Gibco BRL) in 5 µl final volume, starting from the total RNA of 10 equivalent embryos or 15 ng of total tissue RNA. First-strand cDNAs were poly(dG)-tailed using terminal deoxynucleotidyl transferase (20 units; Promega, Madison, WI). PCRs were performed in a total volume of 50 µl using 15 units of Goldstar DNA polymerase (Eurogentec, Seraing, Belgium) for each round of PCR cycles. Samples were incubated at 94°C for 10 min before the two rounds of PCR (15 cycles each; 94°C for 2 min, 63°C for 50 sec, and 72°C for 6 min). PCR products were phenol/chloroform extracted and then purified with the Jet Pure kit (Quantum Appligene, Illkirch, France) in order to eliminate free nucleotides and primers. After quantification by optic density measurement, 400 ng of embryonic cDNA or 1 µg of tissue cDNA were slot-blotted on Hybond N+ nylon membranes (Amersham, Orsay, France). Probes were generated with 20 ng of plasmid-containing insert, labeled by random priming using a Rediprime labeling kit (Amersham). Slot blot membranes were hybridized overnight at 65°C with these probes corresponding to the different inserts (2 x 106 cpm/ml) in hybridization buffer (5x Denhardt, 6x SSC, 0.5% SDS, and 100 µg/ml of salmon sperm DNA) and washed as described above. Membranes were exposed overnight to a phosphorscreen (PhosphorImager, Amersham) and signals were quantified with ImageQuant (Amersham) software. Henrion et al. [26] already showed that expression patterns generated in these conditions reflect the actual transcript evolution.
A pilot experiment was realized by adding increasing amounts of the exogenous "control Neo pa RNA" (Neomycin; Boehringer-Mannheim) to a constant quantity of blastocyst total RNA. Exogenous RNA ranged from 0.05% to 2% of blastocyst mRNA populations. This material was then subjected to a global RT-PCR procedure using identical conditions as those used to generate the spotted cDNA. The amplified cDNAs were slotted onto membranes and hybridized with a radiolabeled Neomycin probe.
Expression patterns in adult tissues are usually analyzed by Northern blotting, which requires 2 µg of poly(A) RNA per lane. Because the majority of the unknown transcripts identified here are absent from adult tissues, a positive control for embryonic expression is mandatory in order to validate the absence of labeling in adult tissues. Because it is impossible to purify 2 µg of embryonic mRNA, we carried out RT-PCR to generate the cDNAs that were slotted onto membranes. Moreover, the same technique was used to generate adult and embryonic samples in order to analyze both with the same sensitivity.
Statistical Analysis
To analyze the patterns of expression of 17 transcripts during preimplantation development, three distinct sets of slot blot membranes (membranes 1, 2, and 3) were performed starting from three distinct batches of embryos at each stage. Probes corresponding to the same clone were independently labeled with 32P and hybridized to one membrane of each set. Hybridization signals were quantified. For most of the 17 genes, 3 repetitions were available, but for 5 genes, only 1 or 2 repetitions could be quantified.
This unbalanced design was analyzed using the software package from Statistical Analysis Systems (Cary, NC). Type III sum of squares was used because of the unbalanced design. We fit a linear model on the expression values transformed by the base 10 logarithm. We assumed and graphically checked that the errors were independent, with equal variance, and normally distributed with a mean of zero. The gene, stage, and membrane effects, and their interactions were tested in the model. ANOVA needs to proceed further in order to determine the details of differences on the stage effect and on the interaction of gene x stage. We used two different approaches; one was to test a set of specific hypotheses (or contrasts), the alternative was to compare all pairs of levels in a factor. If the P value was < 0.05, the test was considered significant.
Screening the Rabbit Genomic DNA Bacterial Artificial Chromosome Library
A rabbit bacterial artificial chromosome (BAC) library was constructed previously [27]. It consisted of 84 480 clones with an average insert size of 100 kilobases (kb) representing a 3-fold coverage of the rabbit genome. The clones were stored in 96-well microliter plates, and the entire collection was organized in a three-dimensional pooling system. The BAC library was screened using primers specific for the 18 inserts with partial or no homology with known sequences. The amplified fragments were between 90 and 208 base pairs (bp) in length. The PCR reactions were performed with Goldstar Taq polymerase (Eurogentec). The cycle parameters used were 94°C for 20 sec and 55°C for 20 sec (35 cycles). The distribution of BACs in this kind of library follows the Poisson law as described by Schibler et al. [28]. We can thus calculate the probability of obtaining a given number of corresponding BACs (addresses) for one insert. Because this BAC library is 3 genomes equivalent, some inserts may be present in several BACs. In addition, because the average size of BAC clones is 100 kb and the size of most eukaryotic genes ranges between 30 and 60 kb, we hypothesized that most genes were contained in a single BAC. Moreover, the probability of having an identical address for two independent inserts is very low (1/30 000), corresponding to the size of the BAC/size of the mammalian genome.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The aim of our study was to identify transcripts synthesized by rabbit embryos, especially those expressed just after ZGA (i.e., the 32- to 64-cell stage) but not present as maternal transcripts at the 1-cell stage. Using SSH we constructed a cDNA library enriched in these transcripts by subtracting 1-cell-stage cDNAs from those that were present at the 32- to 64-cell-stage. We subjected 1000 colonies of this library to differential screening with complex probes that correspond to the cDNAs from 1-cell and 32- to 64-cell-stage embryos (see Materials and Methods). Four hundred colonies were hybridized with at least one probe, and for each, we compared radiolabeled signals after 72 h of exposure. As expected, 90% of the colonies showed a weaker response to the 1-cell-stage probe than to the morula-stage probe.
Insert Sequencing and Identification
Hybridization signals were classified into three groups (Fig. 1): intensities were unchanged between the two conditions (A), intensities were higher at the morula stage (B), or signals were obtained with the morula probe but not with the 1-cell-stage probe (C).
|
The 114 inserts corresponding to group C revealed 50 distinct sequences, 29 of which (25%) were represented once. Of the remainder, 1 insert was represented 21 times, 8 were represented twice, 6 were represented 3 times, 3 were represented 4 times, and 3 were represented 6 times. A search of sequences in the standard international databases revealed that 34 of our 50 sequences (68%) displayed high similarity to known sequences along the entire insert, including two rRNAs. Some sequences corresponded to several contiguous inserts that are homologous to proteins that have already been characterized; for example, ribosomal protein (RP) L7 and cytochrome c oxidase II were represented by three different contiguous inserts; RP S4, RP S6, and ATPase su 8 were represented by two different contiguous inserts; and the 20 remaining proteins or rRNAs were each represented once. The presence of contiguous sequences is explained by use of the SSH method, which incorporates an enzymatic digestion step of all the cDNAs before subtraction. We thus identified 25 already-characterized proteins or rRNA, and 2 sequences that were homologous to proteins of unknown function (clones 5 and 54; see Table 1).
|
Three inserts displayed low identity over their entire sequence (clones 16 and 30) or high homology over short fragments (clone 109) compared with proteins listed in the SwissProt database or with nucleotide sequences listed in databank (Table 2).
|
The sequence of seven cDNAs did not significantly match known gene products (not shown), and nine cDNA inserts showed great similarity to mouse or human ESTs or HTGS (Table 2).
Patterns of Gene Expression During Preimplantation Development
We then analyzed the expression patterns of the isolated transcripts throughout the preimplantation development period in rabbits. We set up a semiquantitative method using PCR-amplified cDNA that we had obtained at each preimplantation stage. This method was validated in a pilot experiment that included increasing the proportions of a reporter exogenous RNA (see Materials and Methods). Quantification of the reporter cDNA shows that signal intensity increased linearly from 0.05% to 1% with the increase in reporter RNA (Fig. 2). The signal still increases for transcripts that range from 1% to 2% of the mRNA population, although probably not linearly. Anyway, very few transcripts in an eukaryotic cell represent more than 1% of the messengers. This semiquantitative method is suitable for studying the expression patterns of the transcripts we isolated by differential screening because complex probes classically hybridize with the cDNA molecules that constitute more than 0.05% to 0.1% of the cDNA population. The transcripts we studied thus probably represented between 0.05% and 1% of mRNA at the morula stage.
|
To quantify preimplantation patterns of expression, we focused on a group of 17 transcripts that represented 13 unknown genes (among 16), and 4 known genes: ribL7 (RPL7), uba80, pros27, and SAMDC.
We measured the expression of these 17 genes in 5 preimplantation stages and with -amanitin treatment. We performed the experiments in triplicate, corresponding to three different sets of membranes, but the expression profile of 5 genes was available only for one or two of these sets. ANOVA (Tables 3 and 4) shows that all sources of variation or effects were significant; the main ones being gene and stage, and the interactions of gene x stage and gene x membrane.
|
|
Because the F-test was significant for the stage effect, we were interested in learning more about the interstage differences and used planned comparisons (contrasts). Considering the data, the test of the null hypothesis (the average expression at 1-cell and 4-cell stages is equal) cannot be rejected (P = 0.1254). When we compared the average expression at the 8- to 16-cell stage with that at the 1- and 4-cell stages or in -amanitin treated embryos we concluded that expression in 1- and 4-cell stages or after -amanitin treatment was different from that in the 8- to 16-cell stage (Table 5). These results point to the efficiency of our subtraction and screening procedures because the isolated genes are clearly expressed from the zygotic genome (Figs. 3 and 4).
|
|
|
Expression at the morula stage significantly differs from that at the 8- to 16-cell stage (P = 0.0029). This probably reflects the dramatic decrease in transcript of some genes (see for example, clones 1, 8, 36, 52, 58, 61, and SAMDC; Fig. 4) at the morula stage.
This global interstage comparison does not detect any significant differences in expression between blastocyst and 8- and 16-cell stages (P = 0.148).
To further analyze the behavior of the different genes at the morula stage, we made all pairwise comparisons on the (gene x stage) interaction. We studied only the 12 genes that had available results of expression on 3 membrane sets. We focused on the differences in average expression (the least-squares means) between the 8- to 16-cell and morula stages for each gene (Table 6).
|
For RibL7, uba 80, clone 5, clone 30, and clone 72, expression at the morula stage does not differ from that at the 8- to 16-cell stage (P > 0.05). Statistical analysis confirms the patterns observed in Figure 4 showing a gradual increase in expression level from the 8- to 16-cell stage onward. For this reason these genes were considered to be "long-term induced."
The opposite is seen in clones 1, 8, 22, and 53, which show significantly lower expression at the morula stage than they do at the 8- to 16-cell stage (P = 0.02, 0.004, 0.001, and 0.04, respectively). Figure 4 shows that SAMDC expression very likely decreases at the morula stage (P = 0.052; at the threshold of statistical significance). Expression of these genes first increases at the 8- to 16-cell stage, then decreases at the morula stage. These genes were considered to be "transiently expressed."
For the two remaining clones (27 and 61), using statistical analysis we could draw no conclusions for the decrease in their expression at the morula stage, due to the high variability of the results obtained on the different sets of membranes. Nevertheless, Figure 4 shows that clone 61 repeatedly exhibited a transient pattern of expression on three distinct membrane sets. A decrease in the expression of clone 27 is probably slower at the morula stage than it is in bona fide, transiently expressed genes, thus resulting in a nonsignificant difference between the 8- to 16-cell and morula stages. However, the difference between the blastocyst and the 8- to 16-cell stages is statistically significant for this clone (see below). Clones 27 and 61 could not be classified as long-term induced genes; rather they are more likely to be transiently expressed.
To further characterize the behavior of transiently expressed genes, we compared the average expression (via least squares means) of each gene between the 8- to 16-cell and blastocyst stages (Table 6). Clones 1, 8, 22, and 27 correspond to genes in which expression is significantly lower at the blastocyst stage than it is at the 8- to 16-cell stage. In clones 53, 61, and SAMDC, the opposite occurs in that expression reincreases at the blastocyst stage.
Characterization of Unknown Genes Expressed During Rabbit Preimplantation Development
Seven clones did not reveal any homology with sequences listed in the databases, 9 clones corresponded to ESTs or HTGS, 2 of which (clones 16 and 30) showed 77% and 47% similarity with known amino acid sequences. Thus we decided to further characterize the corresponding transcripts. We screened a BAC genomic DNA library and we analyzed their expression in adult tissues.
Screening of a BAC Genomic DNA Library
The SSH method may generate contiguous fragments of the same cDNA. To determine the number of new genes that we identified in this study we screened a rabbit genomic DNA BAC library [27] with isolated clones to assign addresses to each unidentified insert. Due to the BAC library properties (see Materials and Methods), we assume that two inserts with the same address most probably belong to the same gene.
We screened the genomic library using PCR with specific primers designed for the 16 unidentified inserts, as well as for clones 5 and 54, which were used as controls. Results are summarized in Table 7.
|
We did not find addresses for clones 43 and 61, which suggests that their genomic counterpart was absent from the BAC library or that it is divided between different BACs. The address of 6 other clones (clones 5, 54, 58, 72, 109, and 128) was not shared by others, which suggests they were held by 6 different genes. We verified that clones 5 and 54 originated different genes from the others as expected. Two addresses were common to clones 16, 30, and 122, whereas BAC 817F1 was common to clones 16 and 122. Therefore, these three clones correspond to a single gene.
The seven remaining clones (1, 8, 22, 27, 36, 52, and 53) form a group and might belong to a multigenic family because they share a common address (see Table 7). Taking into account both the common and nonshared addresses that we found for these seven clones, we estimate that the distribution of these sequences corresponds to at least four different genes that belong to the multigenic family.
In conclusion (see Table 8), we isolated at least nine novel genes that are expressed at the onset of zygotic transcription. This number may increase to 11 if we hypothesize that clones 43 and 61, for which no address was found in the library, represent two distinct genes without any relation to the first 9 considered genes. These two genes are probably distinct because their pattern of expression during preimplantation development varies (see Table 8), but their relationship with the other nine genes remains unknown.
|
Expression Pattern in Adult Rabbit Tissues
Because most of the known genes are expressed ubiquitously, we wondered whether this was also true for the unknown genes. Five of the transiently expressed clones (clones 1, 8, 22, 27, and 61) were used as probes on cDNAs from 10 adult rabbit tissues (ovary, testis, heart, brain, lung, muscle, intestine, kidney, spleen, liver; see Materials and Methods). Complementary DNA from embryonic stages 1 and 8–16 were used as positive controls and were labeled by all the probes that showed the expected differential level of signal between the two stages.
Clones 1 and 22 probes were not able to label any tissue. Clone 8 probe weakly labeled the spleen and the liver (37-fold weaker than the 8- to 16-cell stage embryo) and only weakly labeled testis, lung, and kidney. The clone 27 probe labeled all tissues, but very weakly compared with embryonic controls (between 30- and 160-fold weaker than the 8- to 16-cell stage embryo; Fig. 5). Clone 61 was expressed only in the lung and spleen (data not shown). Clone 5, which was progressively expressed during preimplantation development, was also used as a probe and was expressed mainly in lung tissue. A lower expression was also observed in testis, muscle, spleen, kidney, and liver (data not shown). Clone 26, which encodes RP L34 and is assumed to be ubiquitously transcribed in adult tissues, was used to control cDNA quality. This probe was able to strongly label all tissue cDNAs deposited on the slot blot membranes.
|
In conclusion (see Table 8), at least some novel genes isolated in this study do not encode for ubiquitous transcripts. Among the 4 members of the multigenic family transiently expressed at ZGA, 3 distinct patterns of expression were detected in the 10 adult tissues we tested: no expression (clones 1 and 22), expression restricted to few tissues (clone 8), and ubiquitous but weak expression (clone 27).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Transcription of Genes Encoding Known> Functions at ZGA
Most of the already identified genes we isolated encode proteins or RNAs that work in polypeptide translation or degradation. This transcription at ZGA correlates with an extensive reprogramming of the protein synthesis pattern that occurs when numerous oocyte-specific polypeptides up to the 8-cell stage disappear [32], and when new polypeptides are translated from zygotic transcripts. It also correlates with a great increase in protein content during preimplantation development. The total amount of proteins starts to increase at the late morula stage, then increases rapidly between the late morula and early blastocyst stages (a 2-fold increase in 24 h), and again between the early and late blastocyst stages (a 5-fold increase in the next 24 h). This rise in protein content is correlated with rapid cell division at the late morula/early blastocyst stage, and an increase in the embryo's surface area and volume [29–31].
Two of the genes that we identified, ATPase 6 and SAMDC, may have the potential for use in analyzing the effect of microenvironmental perturbations on the developmental program at ZGA. We found a higher expression of ATPase 6 and SAMDC at ZGA in rabbits. This result is interesting because both these genes are highly expressed at the 2-cell stage (ZGA) in mouse embryos cultured in an oviductal environment that allows development beyond the onset of embryonic transcription. In contrast, they are only slightly expressed in this species when cultured in conditions in which embryo development is blocked at the 2-cell stage [17]. We are presently analyzing the requirement of these gene products for post-ZGA development.
A third gene, SUPT4H, the mammalian homolog of yeast SPT4 [33], may have a role in ZGA regulation itself. In mice, this factor is able to negatively regulate transcription elongation when it is bound to SUPT5H [34, 35]. In rabbits, acquisition of transcriptional control via repressive structuration of chromatin is progressive from the first mitosis to ZGA [5]. Because we showed that zygotic SUPT4H is transcribed from the 8- to 16-cell stage onward, the corresponding protein may be involved in this transcriptional repression. Although this hypothesis is still speculative, it is now open for experimentation.
Unknown Genes Transcribed at ZGA
For the seven sequences that did not match up with known gene products, and the five clones that match up with ESTs or HTGS only, the search for motifs or protein domains did not reveal any significant homology. This may be due to the restricted length of corresponding fragments that may lack a complete motif. However, an alternative hypothesis is that several of these clones correspond to still unknown transcripts. In our study, the nonubiquitous expression of several of these genes may have precluded their previous identification. Moreover, unidentified genes have been described in recent transcriptome analyses of early mammalian embryos [19–21, 36, 37], indicating the high complexity of the expressed genome at this stage of development.
A Developmental Role for Genes Transiently> Expressed at ZGA?
About 20% of the genes we isolated showed transient expression at the 8- to 16-cell stage. Such transient expression has already been described in mice at the 2-cell stage (ZGA) [15, 16, 20, 38, 39], and depending on the study, the proportion of transiently expressed genes varies from 15% [20] to 93% [21] of the genes transcribed at ZGA. Their transcription at ZGA may correspond to a specific developmental role, as is suggested by interspecies conservation of the transient expression pattern of eIF-1A [40]. But it may also result from a still incomplete repressive structure of chromatin soon after acquisition of a permissive transcriptional structure at the beginning of development. In mouse embryos, transient expression at the early 2-cell stage results from the acquisition of a structure that enables access to maternal transcription factors during the first S phase, and from the setting of a repressive structure during the second S phase, which provides the conditions for a more precise spatio-temporal regulation of gene expression [41]. In rabbit embryos, both the first DNA replication and ZGA are sequentially involved in establishing a repressive chromatin state [5]. If transient expression observed in this study is similar to that described in mice, then we must consider the possibility that several replications may be necessary in order to obtain the permissive state (at least for endogenous genes in those species with delayed ZGA), whereas only one replication is sufficient in mice. Again in this scenario, because of their delayed ZGA, rabbit embryos more than mouse embryos offer a better opportunity for investigating the regulatory mechanisms that are responsible for this transcription pattern.
Defining the genes transcribed at the onset of zygotic transcription and determining how their transcription is controlled are goals for molecular embryologists. The differential expression of zygotically expressed genes that we have evidenced immediately after ZGA (corresponding to the morula stage) clearly suggests that early post-ZGA events are also directly involved in regulating embryonic gene expression. Whether genes transcribed at ZGA are functionally related, contain common regulatory sequences that interact with some precise maternal factors, or whether they are spatially clustered in the genome, are still unanswered questions. The clustering of multigenic family members remains relatively rare, but it has been suggested that a regulated expression pattern is clustered on the chromosomal map [21]. Among the transiently expressed genes at ZGA, the identification of several members of a multigenic family and their assignment to precise addresses in the BAC library will be helpful aids in investigating these questions further.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: V. Duranthon, Laboratoire de Biologie du Développement et Biotechnologie, INRA, Domaine de Vilvert, 78352 Jouy en Josas Cedex, France. FAX: 33 01 34 65 26 77; richoux{at}jouy.inra.fr
3 Current address: Laboratoire de Biologie et Génétique du Paludisme, Institut Pasteur, 25 rue du Dr Roux, 75724 Paris Cedex 15, France
Received: 24 August 2001.
First decision: 19 September 2001.
Accepted: 11 July 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-amanitin sensitivity during bovine preattachment development. Mol Reprod Dev 2000 55:152-163[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  V. Duranthon, A. J Watson, and P. Lonergan Preimplantation embryo programming: transcription, epigenetics, and culture environment Reproduction, February 1, 2008; 135(2): 141 - 150. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Shi, F. Dirim, E. Wolf, V. Zakhartchenko, and T. Haaf Methylation Reprogramming and Chromosomal Aneuploidy in In Vivo Fertilized and Cloned Rabbit Preimplantation Embryos Biol Reprod, July 1, 2004; 71(1): 340 - 347. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
