| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Embryo |
a Division of Cell Sciences, School of Biological Sciences, University of Southampton, Southampton SO16 7PX, United Kingdom
b Laboratorio di Technologie della Riproduzione, C.I.Z.-I.S.I.L.S.., 26100, Cremona, Italy
c INRA, Biologie du Developpement, 78352 Jouy en Josas, France
d TEAGASC, Athenry Research Centre, Athenry, Galway, Ireland
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
assisted reproductive technology, early development, embryo, environment
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
+ and ZO-1- variants have been identified in which an 80-amino acid C-terminal domain is differentially spliced [4]. ZO-2 is also alternatively spliced in the C-terminus, either with or without a 36-amino acid ß domain (ZO-2ß± isoforms; [5]). Splice variants such as these may perform distinct roles in TJ activity [4, 6].
During early mouse development, the TJ is constructed in a step-wise assembly process once embryo compaction has occurred. First, the cytoplasmic plaque proteins ZO-1- and rab13 assemble at the 8-cell stage; cingulin plaque protein at the 16-cell stage; followed by the transmembrane protein occludin and the second isoform of ZO-1, ZO-1+, at the 32-cell stage [6–11]. The ZO-2 plaque protein also assembles during the early 8- to 16-cell stage (Nowak et al., unpublished observations). However, most TJ mRNAs are present throughout preimplantation development in the mouse, initially inherited from maternal transcription followed by embryonic transcription from the 2-cell stage onward [3, 11]. Only ZO-1+ is transcribed later from the embryonic genome (16- to 32-cell stage in the mouse [6]). In the bovine, the major activation of the embryonic genome occurs between the 8- to 16-cell stage prior to compaction [12]. However, minor transcriptional activity has been detected as early as the pronuclear stage after in vitro fertilization [12].
It is well established that in vitro produced bovine embryos are retarded in development and often show a reduced degree of compaction before blastocyst formation [13, 14] compared with their in vivo counterparts. In vitro culture has been reported to cause decreased pregnancy rates and long-term effects with higher perinatal death rates in conjunction with malformations and the large offspring syndrome [15]. The immunocytochemistry of TJ formation in bovine embryos appears similar to that of mice [16]. Here, we investigate the pattern of TJ transcription during compaction and blastocyst formation in cattle embryos and evaluate the susceptibility of this genetic program to environmental conditions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
For in vitro production (IVP) in one laboratory, immature oocytes were aspirated from abattoir ovaries and matured in vitro for 20–22 h in tissue culture medium 199 (TCM199) supplemented with 10% fetal calf serum (FCS), LH, and FSH (0.1 IU each; Pergovet, Serono, Italy), insulin-transferrin-sodium selenite (IST) supplement, 10 µg/ml heparin, long-insulin-like growth factor I (IGFI; 100 ng/ml) and long-epidermal growth factor (EGF; 50 ng/ml; all from Sigma, Milan, Italy) at 38.5°C and 5% CO2 in a humidified atmosphere. After fertilization with Percoll-separated sperm in glucose-free synthetic oviduct fluid with amino acids (SOF-aa) supplemented with 6 mg/ml of fatty acid-free BSA (FAF-BSA), 10 mM glycine, 20 µM penicillamine, 10 µM hypotaurine, and 1 µM epinephrine for 19 h, presumptive zygotes were denuded of the cumulus cells and cultured in 20-µl drops of SOF-aa with 16 mg/ml of FAF-BSA and 10 mM glycine in 5% CO2, 5% O2, and 90% N2 in a humidified atmosphere at 39°C (experiments 1, 2, 4, and 5). Half the medium was changed on Days 4 and 6 of culture. On Day 6, the discarded medium was replaced with TCM199 supplemented with the same amounts of BSA and glycine.
For cryopreservation, a conventional ethylene-glycol-based freezing protocol was used. Morulae were incubated in Hepes-buffered SOF medium supplemented with 6 mg/ml BSA and 1.5 M ethylene glycol for 20 min, and loaded into 0.25-ml straws. Using a freezing machine (Biocool, FTS System Inc., Stone Ridge, NY), the embryos were held at -6°C for 5 min, seeded, held for another 5 min, and then cooled down to -32°C at 0.5°C/min before plunging them into liquid nitrogen.
Alternatively, in another laboratory (experiments 3 and 4), embryos were generated in vitro in a complex co-culture system as described by Menck et al. [17]. Briefly, groups of 40–50 in vitro matured oocytes were fertilized in modified Tyrode medium (TALP) supplemented with 10 µg/ml heparin for 18 h at 39°C with swim-up separated sperm at a final concentration of 1 x 106 cells/ml. Denuded zygotes were then cultured in microdrops of B2 medium with 10% FCS on Vero cell monolayers.
Mature oocytes were scored according to the presence of a polar body after in vitro maturation. On the morning of Day 4 after in vitro fertilization, 8- to 16-cell cleavage stages were collected. Compaction grade was determined on the morning of Day 5 (early and long-compacting embryos) or Day 6/7 (late and short-compacting embryos) after fertilization, and blastocyst formation was determined on the morning of the following day, respectively (Day 6/7/8).
Recovery of in vivo produced embryos was carried out as described by Morris et al. [18]. Briefly, Hereford-cross heifers were superovulated with an i.m. administration of 1500 IU of eCG (Folligon, Intervet UK Ltd., Cambridge, U.K.) during the midluteal phase of the estrous cycle (Days 10–14) and 500 µg cloprostenol (Estrumate, Coopers Animal Health Ltd., Berkhamsted, U.K.) 48 h later to induce luteolysis. Following administration of cloprostenol, the heifers were continuously checked for overt signs of estrus and those observed in standing estrus were artificially inseminated by one operator using semen from one sire.
On Days 4 to 8 after insemination, embryos were flushed from oviducts, uteri, or both to obtain 8- to 16-cell embryos, morulae, blastocysts, and expanded blastocysts. The composition of the flushing medium was 139 mM NaCl, 2.7 mM KCl, 0.89 mM CaCl2·2H2O, 1.47 mM KH2PO4, 0.49 mM MgCl2·6H2O, 7.46 mM Na2HPO4·2H20, 1 mM glucose, 0.5 mM sodium pyruvate, and 0.1% (w/v) polyvinyl alcohol (PVA) pH 7.3.
All oocytes and embryos after in vitro culture or in vivo embryos immediately following recovery (only grade 1 and 2 embryos) were washed five times in 10 mM PBS containing 1 mg/ml PVA, snap-frozen individually in a minimal amount of medium (2 to 5 µl), and stored at -80°C until mRNA extraction. Blastocysts selected for mRNA expression analysis were morphologically normal and had a clear cavity, a distinguishable inner cell mass, and an intact zona pellucida.
Isolation of RNA
Total RNA was extracted from bovine liver using a commercial kit according to the manufacturer's instructions (RNeasy Midi Kit, Qiagen, Crawley, U.K.) and subsequently treated with 1 U RQ1 DNase (Promega, Southampton, U.K.) for 60 min at 37°C. Poly(A)+ RNA was isolated from single bovine oocytes/embryos using magnetic oligo(dT) beads (Dynabeads mRNA DIRECT Kit, Dynal, Oslo, Norway) according to the manufacturer's instructions with minor modifications [19]. Briefly, 150 µl of lysis-binding buffer (100 mM Tris-HCl pH 8.0, 500 mM LiCl, 10 mM EDTA, 1% w/v lithium dodecylsulfate (LiDS), 5 mM dithiothreitol [DTT]) was added to the frozen embryo, mixed, and incubated at room temperature for 10 min. Subsequently, 10 µl of washed Dynabeads Oligo(dT)25 in lysis-binding buffer and 1 pg/µl luciferase mRNA (Promega) as external standard were added in semiquantification experiments. For hybridization, beads, lysate, and standard RNA were incubated for 10 min at room temperature on a roller shaker, centrifuged at 12 000 x g for 15 sec, and placed in a chilled MPC-P-12 magnet. The beads with the bound poly(A)+RNA were washed twice in 100 µl wash buffer A (10 mM Tris-HCl pH 8.0, 0.15 M LiCl, 1 mM EDTA, 0.1% w/v LiDS), and three times with 100 µl of washing buffer B (10 mM Tris-HCl pH 8.0, 0.15 M LiCl, 1 mM EDTA). The poly(A)+RNA was then eluted from the beads by incubation in 10 µl of sterile water for 2.5 min at 65°C and separated from the beads on a chilled magnet.
Reverse Transcription
From isolated RNA, 80% was reverse transcribed into cDNA in a total reaction volume of 20 µl consisting of 1x reverse transcriptase (RT) buffer (50 mM Tris-HCl pH 8.3, 75 mM KCl, 3 mM MgCl2, and 100 µM DTT, Gibco BRL, Paisley, UK), 20 U RNase inhibitor (Perkin-Elmer, Foster City, CA), 5 µM random hexamer primers (Perkin-Elmer), 500 µM of each dNTP, and 200 U Superscript II RNase H- RT (Gibco BRL). To control for contamination with genomic DNA, the remaining RNA (20%) was used in a similar reaction omitting the RT.
Polymerase Chain Reaction
Gene-specific fragments were amplified in polymerase chain reaction (PCR) mixtures consisting of 1x PCR buffer (20 mM Tris-HCl pH 8.4, 50 mM KCl), 1.5 mM MgCl2, 200 µM of each dNTP (all from Gibco BRL), 400 µM of each sequence specific primer, and cDNA at the concentrations determined during the optimization experiments to produce reliable amplification within the linear range (Table 1). All reactions were overlayed with 50 µl of mineral oil (Sigma, Dorset, U.K.). After initial denaturation at 95°C for 3 min, 2.5 U of Taq DNA polymerase (Gibco BRL) were added during the hot start at 72°C for 30 sec. The cycling program on a PTC-225 thermocycler (MJ Research, Watertown, MA) consisted of denaturing at 95°C for 30 sec, specific primer annealing temperature for 30 sec, and template extension of 72°C for 30 sec. The number of cycles required to amplify each gene-specific fragment within the linear range was determined in initial optimization experiments (Table 1). A 10-min final extension step at 72°C was followed by holding at 10°C. From reactions, 10 µl of PCR product were added to 2 µl of 6x loading buffer (15% Ficoll, 0.02% bromophenol blue; Sigma) and 10 µl of this was subjected to electrophoresis in TE buffer on a 2% agarose gel containing 1 µg/ml ethidium bromide. To estimate fragment sizes, a molecular weight marker (100 base pairs, Gibco BRL) was run alongside.
|
To ensure the absence of contaminating DNA, each reverse transcribed cDNA sample was tested by omitting RT during the RT reaction. Occludin primers located on the same exon were used as a control for DNA contamination during mRNA extraction and RT-PCR. All other bovine-specific primers were located on different exons and resulted in larger amplification products when employed to DNA (data not shown). As an additional negative control for cDNA carryover, a PCR containing all reagents and each primer pair but no cDNA template was run alongside for each experiment (reagent control). As positive control for the RT-PCR, approximately 100 ng of total RNA from bovine liver was amplified in parallel.
Production of Bovine-Specific Gene Fragments and Primers for PCR
Initially, heterologous primer pairs were designed according to human, canine, or bovine TJ sequences as published in GenBank and obtained from Oswel (Southampton, U.K.). With these primers, bovine-specific gene fragments were amplified from total liver RNA using nonstringent RT-PCR conditions and suboptimal annealing temperatures. Amplification products of the correct fragment sizes were sequenced (ABI prism 377, ABI, Warrington, U.K.) and compared with the GenBank database for homology (Table 2). Bovine-specific primer pairs were then designed from those sequences for the TJ constituents pan ZO-1, ZO-1+, occludin, JAM, ZO-2ß+, and pan ZO-2 (Table 1). RT-PCR amplification products from embryo material using the bovine-specific primers were sequenced and showed 96%–99% identity with the tissue derived RT-PCR products.
|
Semiquantification of the RT-PCR
After digitization of the gel image using a digitized camera system (Alpha Imager 1220, Alpha Innotech Corporation, San Leandro, CA) and a constant superimposed overlay produced with the AlphaEase software (Alpha Innotech Corporation, San Leandro, CA), integrated density values (IDVs) were generated, calculated as the sum of all pixel values after background correction. A control value from a region of no product was subtracted from the density data if it was found to be greater than zero. The abundance of a given transcript was determined as the ratio of the IDVs from the luciferase standard amplification product to the gene-specific amplification product. This approach can be used to compare the relative abundance of one mRNA among different samples, but not the absolute amount of one mRNA compared with another. For each embryo, the total amount of TJ transcripts was calculated by summing the relative abundance values for each single transcript. When a single transcript could not be quantified due to contamination of the reagent control (see above), this individual embryo was excluded from total TJ transcript analysis. The different embryo categories compared with each other within one experiment were always run in parallel in at least six replicate experimental runs to minimize variation due to day-to-day RT-PCR variability. Experimental runs were entirely excluded from statistical analysis when two or more embryo categories could not be examined for more than one transcript.
Statistical Analysis
Data were analyzed using the SigmaStat 2.0.3 (Jandel Scientific, San Rafael, CA) software package. After testing for normal distribution and equal variance, either a t-test (two group comparison) or ANOVAs (multiple group comparison and analysis for main effects) were performed. A two-way ANOVA or a two-way ANOVA on ranks was used where appropriate, followed by a Tukey test or a Dunns test to identify the significantly different treatment groups. Differences of P < 0.05 were considered to be significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the simple SOF-aa based IVP system, in vitro maturation as determined by polar body extrusion exceeded 80%, and 76.9% oocytes cleaved (412/536). On Day 5, 17% (91) were compact morulae (CM5), whereas 48.7% (261) remained noncompacted on the morning of Day 6 (non-CM). On the morning of Day 7, 14.2% (76; CM5) and 3.7% (20; non-CM) developed to blastocysts from the different compaction groups. Delayed blastocyst formation on Day 8 occurred for 1.5% (8; CM5) and 6.2% (33; non-CM), respectively. Overall, 35.5% embryos developed to blastocysts by Day 8. On average, good quality Day 5 morulae contained 39.7 ± 2.4 (n = 25) cells and their Day 7 blastocysts had 151.2 ± 5.5 cells (n = 21), whereas noncompact Day 6 morulae contained 32.9 ± 2.8 (n = 25) cells and their Day 7 blastocysts contained 75.2 ± 4.7 (n = 25) cells, significantly less (P < 0.01) than blastocysts derived from CM5 morulae. Using a conventional freezing protocol based on 1.5 M ethylene glycol, freeze/thaw-survival or hatching in vitro of Day 7 blastocysts derived from non-CM morulae (n = 32) was significantly decreased compared with blastocysts derived from CM5 morulae (n = 40) during 72 h after thawing (CM5 blastocysts, 97.5%; 95%, and 95% viable, respectively, 24, 48, or 72 h after thawing; and 75% hatched at 72 h after thawing; non-CM-blastocysts, 18.7%, 9.4%, and 9.4% viable, respectively, 24, 48, or 72 h after thawing, and 6.3% hatched at 72 h after thawing).
In the complex B2-co-culture based IVP system employed in experiments (experiments 3 and 4), 95.4% (166/174) of the embryos cleaved on Day 2. On Day 8, 72.9% (127/174) had reached the blastocyst stage. Upon transfer to synchronized recipients, the 3-mo pregnancy rate was 58% (15/26 [17]).
Similar Ontogeny of TJ Transcription During Preimplantation Development In Vitro and In Vivo (Experiment 1)
Immature germinal vesicle (GV) and in vitro matured metaphase II (MII) oocytes, 8- and 16-cell embryos, morulae, and blastocysts generated in vitro in SOF-aa and in vivo were analyzed for the presence or absence of TJ mRNA expression (Fig. 1A–C). The plaque proteins pan ZO-1, ZO-1+, ZO-2ß+, and pan ZO-2 showed similar mRNA expression patterns in embryos derived in vitro and in vivo. The percentage of embryos expressing these transcripts from the maternal genome decreased from the oocyte stage to the 8- and 16-cell stage before increasing again by the morula stage, indicating embryonic genome origin. In contrast, occludin expression remained stable in in vitro derived embryos between the 8- and 16-cell stage (80% expressing) whereas in in vivo generated embryos, maternal occludin transcripts had degraded by the 16-cell stage. In vivo, embryonic transcription increased to the same proportions as for in vitro embryos by the morula and blastocyst stage. In contrast to the other TJ transcripts, maternal JAM transcript was found only in a few GV and MII oocytes, but it was present in every embryo studied from the 8-cell stage onward.
|
Validity of the Semiquantitative RT-PCR Assay
In the semiquantitative approaches, for each transcript the proportion of isolated mRNA from a single embryo necessary to achieve good amplification with gene-specific primers at 40 PCR cycles was determined. Using this proportion and increments of lower PCR cycles, a cycle number for PCR within the linear range for each transcript was selected. The reproducibility of amplification for the exogenous standard luciferase mRNA and for the endogenous transcript JAM from a single embryo were determined as coefficients of variation (CVs) of 7.7 for luciferase (22 cycles, n = 4; Fig. 2A) and 4.0 (34 cycles, n = 4; Fig. 2B). This was constant and low compared with previous studies (6%–46% [25]) and allowed us to detect reliably less than 1.5-fold differences between embryo categories when at least six embryos were examined in each category. Six different endogenous TJ transcripts and the external luciferase standard were amplified reliably from each single embryo (Fig. 2C).
|
Stable Maternal Inheritance of TJ Transcripts During Oocyte Maturation In Vitro (Experiment 2)
Using PCR conditions within the linear amplification range, which gave highly reproducible results in morulae and blastocysts (see above), expression levels of pan ZO-1, ZO-1+, and occludin were high in GV and MII oocytes compared with blastocysts (see below and
Figs. 4–6
). Levels of ZO-2ß+ and pan ZO-2 were lower in oocytes than in average blastocysts, and JAM expression levels were barely detectable. This remained constant during oocyte maturation for each transcript and for the total amount of TJ transcripts (Fig. 3A–C).
|
|
|
|
Effect of Timing of Development In Vitro on the Relative mRNA Expression Levels of TJ Transcripts (Experiment 3)
To determine the effects of the timing of morula and blastocyst formation on TJ mRNA expression levels, the relative abundance of pan ZO-1, ZO-1+, occludin, JAM, ZO-2ß+, and pan ZO-2 transcripts was determined in single early (Day 5 morulae, Day 6 blastocysts) or late (Day 7 morulae, Day 8 blastocysts) developing embryos generated in vitro in a complex B2 co-culture based system (Fig. 4A–K). Overall, the relative abundance of pan ZO-1, ZO-1+, and occludin increased significantly (P < 0.01–0.05) upon blastocyst formation in both early and late developing embryos. However, late developing blastocysts had significantly more pan ZO-1 transcript than early developing blastocysts. The relative expression levels of JAM, ZO-2ß+, and pan ZO-2 transcription were not altered significantly by timing of development or during morula to blastocyst transition. However, although the relative abundance of JAM, ZO-2ß+, and pan ZO-2 was lowest in Day 7 morulae, which represent very retarded embryos, ZO-1+ relative expression levels increased constantly with the day of development irrespective of the embryo stage (Fig. 4B). The total amount of TJ transcripts increased significantly (P < 0.05) during blastocyst formation within the later developing embryos.
TJ Transcript Levels in Blastocysts in Relation to Compaction Period and In Vitro Culture System (Experiment 4)
After slightly different in vitro maturation and fertilization steps (see "Materials and Methods"), individual in vitro produced embryos derived from either a simple (mSOFaaBSA), low-oxygen culture system or a complex (B2 with serum) co-culture system with high oxygen were scored according to the length of the compaction period: short-compacting embryos were Day 6 noncompacted morulae (non-CM) and Day 7 blastocysts derived from them, long-compacting embryos were Day 5 compact morulae (CM5) and Day 7 blastocysts derived from them. Single blastocysts that developed in both IVP systems in different laboratories at Day 7 after a long or a short compaction period were analyzed for TJ transcription levels (Fig. 5A–G) after accounting for the different origin of the embryos (two-way ANOVA). Relative expression levels of all TJ transcripts did not differ significantly in the two IVP systems. However, the total amount of TJ transcripts was increased significantly (P = 0.04) in blastocysts derived from long-compacting morulae compared with short-compacting ones, in particular within the complex co-culture based IVP system. Although there was a tendency for all transcript levels to be increased after a longer compaction period in both IVP systems, this was significant (P = 0.012) only for ZO-2ß+ transcript levels, in particular in the complex co-culture based IVP system.
TJ Transcript Levels in Relation to Developmental Stage, Compaction Period, and In Vitro or In Vivo Origin (Experiment 5)
To determine whether the length of the compaction period before blastocyst formation influences mRNA expression, the relative abundance of TJ transcripts was determined in in vitro and in vivo derived morulae and blastocysts (Fig. 6A–G). In vitro produced embryos generated in the simple (mSOF-aa BSA) culture system were scored as defined in the previous section: short-compacting embryos as non-CM morulae (Fig. 6H) and blastocysts (Fig. 6I), long-compacting embryos as CM5 morulae (Fig. 6J) and blastocysts (Fig. 6K). These two in vitro groups were compared with in vivo derived compacted Day 5 morulae (Fig. 6L) and Day 6 blastocysts (Fig. 6M).
The relative expression level of several transcripts increased significantly during the morula to blastocyst transition in in vitro and in vivo generated embryos. Overall, the TJ mRNA level increased 6-fold in short-compaction embryos (P < 0.001) and 2.5-fold in in vivo derived embryos (P = 0.04; Fig. 6G). The increase was 2.5-fold to 5-fold for pan ZO-1 and 5-fold to 10-fold for ZO-1+ (P < 0.001). However, for other transcripts (occludin, ZO-2ß+, and pan ZO-2), a significant increase in transcript levels between morula and blastocyst stages was apparent only in the short-compaction in vitro embryo group (Fig. 6C, E, and F). The JAM mRNA level was significantly increased in long-compaction morulae compared with short-compaction morulae (Fig. 6D). The level of pan ZO-1 mRNA was 1.5-fold to 3-fold (P = 0.05) higher in in vivo generated morulae and blastocysts than in short-compaction in vitro ones. Thus, transcript expression in short-compaction in vitro embryos differed more markedly from the levels found in in vivo embryos than did those of long-compaction in vitro embryos. However, most TJ mRNA levels were highest in in vivo derived morulae (3-fold to 4-fold higher) or blastocysts (0.5-fold to 2-fold higher) compared with in vitro counterparts.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Most TJ mRNAs in cattle embryos followed the most common maternal and embryonic transcription pattern [26], which is similar to that of the mouse [3, 11]. This expression pattern was not affected by in vitro conditions for four transcripts, namely, pan ZO-1, ZO-1+, pan ZO-2, and ZO-2ß+. However, whereas mRNA expression of occludin followed a similar pattern in in vivo derived embryos, transcripts persisted in in vitro derived ones during the 8- to 16-cell stage. This may suggest a change in mRNA stability, and/or a turnover due to in vitro conditions. Different in vitro conditions can affect the duration of the cell cycle or alter cell cycle stages, which may cause persistence of maternally derived factors [14]. A similar phenomenon has been reported for the specific leukemia inhibitory factor (LIF)-LIF-receptor system in in vitro generated bovine blastocysts [19], a cytokine signaling system that regulates cell differentiation in the mouse [27]. Alternatively, embryonic transcription of occludin may be initiated prematurely in vitro, therefore masking degradation of maternally inherited message. A similar, very early initiation of presumably embryonic mRNA expression was found in the present study for JAM, which was not affected by in vitro or in vivo conditions and previously for LIF in a manner that was dependent on oocyte batch [19]. Taken together, these findings add to previous data showing gross perturbations of mRNA expression patterns in in vitro derived bovine embryos [26].
Employing our semiquantitative assay, a dramatic increase of total TJ transcripts was observed, in particular of pan ZO-1, ZO-1+, and occludin, during the transition from morulae to blastocysts both in vivo and in vitro. This was also evident in late developing blastocysts in vitro (Day 8). This suggests a stage-dependent rather than a time-dependent up-regulation of embryonic transcription of particular TJ genes just prior to cavitation and cannot be entirely explained by increases in total cell numbers. Whether the relative allocation to the two cell phenotypes, ICM or TE, could be responsible, remains to be clarified.
When blastocysts developed after only a short compaction period in vitro, total TJ mRNA expression levels were decreased under complex co-culture conditions and mainly due to decreased ZO-2ß+ transcripts. This suggests that the length of the compaction period may be critical for providing sufficient time for the production of enough TJ mRNAs necessary for cavitation dependent upon the culture system employed. A shorter compaction period has been reported previously in fast-developing embryos generated in more complex conditions in vitro as opposed to in vivo counterparts [28, 29]. This coincided with perturbed allocations to ICM and TE, which was speculated to be one reason for a loss of viability of in vitro generated embryos upon transfer to recipients [28, 29]. Under simple culture conditions, TJ expression levels were consistent in the present study within comparative groups of embryos in experiments 4 and 5 (i.e., Day 7 blastocysts after short or long compaction). This shows that specific and more complex culture conditions, in particular the presence of serum and co-culture with somatic cells, can multiply morphological differences in relation to relative TJ mRNA levels, which remain less pronounced under simpler culture conditions [30]. However, it cannot be excluded that differences in starting material or in vitro maturation and fertilization between the two different laboratories contributed to this difference. Alternatively, stress adaptations of the preimplantation embryo in suboptimal conditions may be responsible. Recent reports of increased mRNA levels of stress indicators (e.g., Hsp70.1, Mn SOD) in embryos generated in the presence of serum support this idea [31, 32]. Ultrastructural analysis of bovine blastocysts generated in vivo or in vitro in the presence of serum showed reduced junctional contacts in the in vitro generated embryos that were also more susceptible to cryoprotectant and cryopreservation [33]. Together with the low cryopreservation survival rate of short-compacting as opposed to long-compacting blastocysts examined in the present study, this may suggest an insufficient function of junctional components (e.g., ZO-2ß+). However, although a ZO-2ß+ isoform has been reported in several species (humans, dogs, chickens), to date, a detailed function of this particular isoform remains unknown. This may also be attributable to the unconserved localization of the ß motif within the transcript [34].
Morulae and blastocysts with a short compaction period in vitro also expressed less pan ZO-1 compared with their in vivo counterparts. Because ZO-1+ mRNA was not affected, this suggests that the ZO-1- isoform was more sensitive to in vitro culture and compaction length than ZO-1+. In mice, ZO-1- is transcribed throughout cleavage and the protein starts to assemble at the membrane during compaction, whereas ZO-1+ is transcribed, translated, and undergoes membrane assembly within a very short period of time just prior to cavitation [6]. It remains unknown whether ZO-1- mRNA is constantly transcribed de novo during cleavage and cavitation or is very stable compared with ZO-1+ mRNA. In the latter, ZO-1- mRNA may be exposed to suboptimal conditions in vitro for a longer period than ZO-1+, which physiologically does not require a long time for expression regulation. Therefore, ZO-1- transcript may be more susceptible in retarded and short-compacting in vitro generated bovine embryos compared with ZO-1+. A change in mRNA stability may also explain the conflicting reports of detection [30] or failure of detection of pan ZO-1 mRNAs in bovine embryos when generated in vitro in the presence of serum and co-culture while the protein was found [16]. In addition, long-compacting morulae derived in vitro possessed significantly more JAM transcripts than short-compacting ones. Long-compacting morulae in vitro had JAM mRNA levels similar to their blastocysts despite more than three times higher cell numbers in blastocysts. Short-compacting or in vivo counterparts still showed an increase of JAM mRNA levels during cavitation. This suggests that JAM mRNAs accumulate in morulae over time independently of cell number and cavitation. Whether this may be due to an increase in de novo transcription or a decrease in transcript turnover rate remains unknown.
Taken together, our results indicate that embryos that showed a detectable and well-formed compaction period in vitro are of similar quality to their in vivo counterparts. Starting from the same amount of maternal message, in vivo and in vitro development differ most during the critical period of the major maternal zygotic transition before a dramatic increase of TJ mRNAs occurs upon blastocyst formation. Failure to compact in vitro results in significant reduction of specific transcript levels, in a manner dependent on culture conditions, which may contribute to reduced viability. In conclusion, we show that TJ mRNA expression levels are sensitive to environmental conditions that may influence the developmental potential of bovine blastocysts.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Judith J. Eckert, Division of Cell Sciences, School of Biological Sciences, University of Southampton, Bassett Crescent East, Southampton SO16 7PX, United Kingdom; jje{at}soton.ac.uk
3 These two authors contributed equally to this study
Received: 6 August 2002.
First decision: 28 August 2002.
Accepted: 30 October 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
+ isoform. Development 1997 124:2027-2037[Abstract]This article has been cited by other articles:
|
|
 |
|
  A. K Singh, Y. Jiang, S. Gupta, and E. Benlhabib Effects of chronic ethanol drinking on the blood brain barrier and ensuing neuronal toxicity in alcohol-preferring rats subjected to intraperitoneal LPS injection Alcohol Alcohol., September 1, 2007; 42(5): 385 - 399. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Y. Kwong, D. J Miller, E. Ursell, A. E Wild, A. P Wilkins, C. Osmond, F. W Anthony, and T. P Fleming Imprinted gene expression in the rat embryo-fetal axis is altered in response to periconceptional maternal low protein diet. Reproduction, August 1, 2006; 132(2): 265 - 277. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. C. Thomas, B. Sheth, J. J. Eckert, G. Bazzoni, E. Dejana, and T. P. Fleming Contribution of JAM-1 to epithelial differentiation and tight-junction biogenesis in the mouse preimplantation embryo J. Cell Sci., November 1, 2004; 117(23): 5599 - 5608. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. P. Fleming, W. Y. Kwong, R. Porter, E. Ursell, I. Fesenko, A. Wilkins, D. J. Miller, A. J. Watkins, and J. J. Eckert The Embryo and Its Future Biol Reprod, October 1, 2004; 71(4): 1046 - 1054. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
