1 Department of Obstetrics & Gynecology, Yale University Medical School, 333 Cedar Street, New Haven, CT 06520, 2 Reprogenetics LCC, 101 Old Short Hills Road, Suite 501, West Orange, NJ 07052, 3 Tyho-Galileo Research Laboratories LLC, 101 Old Short Hills Rd, Suite 501, West Orange, NJ 07052, 4 Division of Reproductive Endocrinology, Department of Obstetrics and Gynecology, Mayo Clinic, Rochester, Minnesota 55905 and 5 Department of Obstetrics and Gynaecology, University College London, 8696 Chenies Mews, London WC1E 6HX, UK
6 To whom correspondence should be addressed at: Department of Obstetrics & Gynecology, Yale University Medical School, 333 Cedar Street, New Haven, CT 06520, USA. Email: dagan.wells{at}yale.edu
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: apoptosis/cell cycle checkpoint/IVF/PGD/RTPCR
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Many human preimplantation embryos generated using assisted reproductive techniques fail to form a viable pregnancy; only 25% implant following transfer 2 days after IVF. It is likely that a high rate of loss also exists among naturally conceived embryos, as evidenced by the high prevalence of early pregnancy failure in women (Short, 1979; Edmonds et al., 1982
). Given the complex and critical nature of the processes occurring at the preimplantation stage, it is perhaps unsurprising that many embryos fail to negotiate this challenging phase of development. Knowledge of the genes expressed at early stages should highlight the pathways that must be activated in order to produce an embryo capable of successfully implanting and initiating a clinically viable pregnancy.
As well as the scientific importance of studying gene activity in early human embryos, there is growing interest in how information on gene expression could be clinically applied in order to improve the success rates of assisted reproductive techniques. Indeed it is conceivable that particular patterns of gene expression may be indicative of embryo viability per se. If such patterns could be assessed prior to embryo transfer they could assist embryologists and clinicians in deciding which embryos to transfer to the uterus. Current evaluations of embryo viability are based on morphology, in some cases supplemented with information on chromosomal status via preimplantation genetic diagnosis (PGD) (e.g. Munné et al., 1993, 2003
; Verlinsky et al., 1996
; Gianaroli et al., 1997
) or non-invasive assessment of metabolic activity (Gardner et al., 2001
).
We have set out to examine the expression of nine genes throughout the preimplantation phase of development. The genes tested have significance in a range of important cellular processes, including cell cycle regulation, DNA repair, apoptosis, maintenance of accurate chromosomal segregation and construction of the cytoskeleton. The number of transcripts derived from each gene was assessed using real-time RTPCR, a highly accurate method for quantification of nucleic acids. In this way, we aimed to provide an indication of each gene's activity during different stages of preimplantation development.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Embryos from The Institute for Reproductive Medicine and Science of Saint Barnabas were classified non-viable, following routine morphological assessment on day 3 post-fertilization (based on rate of cleavage, percentage of embryo volume occupied by fragments and the presence of various abnormal morphological features). The embryos from the Mayo Clinic had been frozen at the pronuclear stage with the intention of later therapeutic use at pronucleate stage. These embryos were morphologically normal at the time of freezing and considered to be potentially viable (Damario et al., 1999). After being donated for research, the cryopreserved embryos were thawed and allowed to develop to designated preimplantation stages before analysis.
Efforts were made to ensure that RNA samples from oocytes and embryos were not contaminated with nucleic acids derived from sperm or cumulus cells attached to the zona pellucida. This was achieved by removing the zona pellucida using a brief immersion in acid Tyrode's solution (Sigma, USA). Ooctyes/embryos were then washed through three droplets of phosphate-buffered saline supplemented with 0.1% w/v polyvinyl alcohol and 0.3 IU/µl RNasin (Promega, USA). Removal of the zona pellucida and washing required <1 min per sample.
After washing, samples were transferred to a microfuge tube in 1 µl of fluid to which 100 µl of denaturing solution (Micro RNA Isolation kit; Stratagene, USA) and 0.72 µl of
-mercaptoethanol was added. Samples were then stored immediately at 80 °C.
RNA extraction
Prior to RNA isolation, each sample was spiked with 106 copies of an RNA transcript derived from a plasmid (pw109, PerkinElmer GeneAmp RNA PCR kit). After RNA extraction the amount of this artificial RNA in each sample was compared to a control. The control consisted of 106 copies of pw109 RNA that had been subjected to RTPCR directly, without passing through an RNA extraction. The extent of material loss during RNA isolation was assessed by comparison to this control.
To each sample (stored as described above) 10 µl of 2 mol/l sodium acetate (pH 4.0), 100 µl of water-saturated acid phenol and 30 µl of chloroform:isoamyl alcohol were added (Micro RNA Isolation kit; Stratagene). The samples were mixed thoroughly then centrifuged for 5 min (16 000 g). The aqueous layers were transferred to fresh 0.5 ml tubes and 1 µl of glycogen (20 mg/ml; Roche, USA) and 100 µl of isopropanol were added. The resultant solutions were mixed and then centrifuged for 45 min (16 000 g). The supernatant was discarded and the pellet washed with 200 µl of 75% ethanol; centrifuged for 5 min (16 000 g); air-dried; and then resuspended in 5.8 µl of diethyl pyrocarbonate-treated water. To eliminate residual genomic DNA from the RNA sample, 0.725 µl of 10x Amplification Grade DNase I buffer and 0.725 IU DNase I (Invitrogen, USA) was added. Digestion proceeded at room temperature for 15 min and was then stopped by adding 1 µl of 25 mmol/l EDTA and heating for 10 min at 65 °C.
Reverse transcription
Reverse transcription was achieved by adding the following to each RNA sample: 0.2 µl dithiothreitol (0.1 mol/l); 1.5 µl oligo dT (50 µmol/l; RNA PCR kit; Perkin Elmer, USA); 1.05 µl RNAse inhibitor (20 IU/µl, PE RNA PCR kit); 4 µl 25 mmol/l MgCl2; 2 µl 10x PCR buffer (PE RNA PCR kit); 4 µl dNTP (2 mmol/l dATP, dCTP, dGTP, dTTP). The total reaction volume was 20 µl.
Reactions were heated to 70 °C for 6 min and then placed on ice. 0.5 µl Moloney murine leukaemia virus reverse transcriptase (PE RNA PCR kit) was then added and reactions incubated at 37 °C for 1 h. Reverse transcription was terminated by heating to 95 °C for 5 min and the resultant complementary DNA (cDNA) samples were stored at 80 °C until required.
Real-time quantitative PCR
To maximize accuracy each sample was tested three times and amplified simultaneously with four different concentration standards (also run in triplicate) and a negative control composed of reaction mixture with no cDNA added. Concentration standards were generated by amplifying the transcript of interest, performing agarose gel electrophoresis, purifying the band corresponding to the amplified transcript using a QIAquick gel extraction kit (QIAGEN, Inc., USA), and then deducing the number of cDNA fragments per microlitre based on the molecular weight of the amplicon and the measured DNA concentration (calculated from absorbance at 260 nm).
Reaction mixtures were assembled in a 30 µl volume containing 2 µl of cDNA, 3 µl of 10 x buffer (30 mmol/l MgCl2; Idaho Technology Inc., USA), 0.5 µmol/l of each primer, 0.2 mmol/l dNTP (Promega, USA), SYBR green DNA stain (1 µl of a 1/4000 dilution of concentrated stock; Molecular Probes, USA), and 1.88 IU Taq polymerase (inactivated by addition of TaqStart antibody; Clontech, USA). Seven microlitres of this reaction mixture was pipetted into each of three capillary tubes. The tubes were sealed and then subjected to thermal cycling using a LightCyclerTM real-time PCR machine (Idaho Technology, Inc.).
Thermal cycling involved heating at 96 °C for 1 min to denature the cDNA and activate the Taq polymerase. This was followed by 4050 cycles of denaturation for 0 s at 95 °C, annealing of primers at 5060 °C (depending on primers) for 0 s and extension at 72 °C for 1015 s. Fluorescence data were acquired during an additional step at 3 °C below the product Tm for 2 s. Primers were designed such that the 3' end of each gene was targeted for amplification, thus minimizing bias introduced by reverse transcription using poly dT oligonucleotides. Product identity was confirmed by ethidium bromide-stained 2% agarose gel electrophoresis and verified by sequencing.
Quantification
Software supplied with the LightCyclerTM allowed analysis of data concerning PCR product accumulation during thermal cycling. A standard curve was generated by amplifying concentration standards containing a known quantity of amplicons (see above). Fluorescence was acquired at each cycle in order to determine the threshold cycle or the cycle during the log-linear phase of the reaction at which fluorescence rises above background for each sample (i.e. amplified product becomes detectable). The LightcyclerTM quantification software generates a best-fit line and determines unknown concentrations by interpolating the noise-band intercept of an unknown sample against the standard curve of known concentrations (Wittwer et al., 1997; Steuerwald et al., 2000
).
Hierarchical cluster analysis
Hierarchical cluster analysis was performed on all 46 samples that received a complete analysis of all nine genes. For this purpose we employed Cluster and TreeView, programs created by Michael Eisen (http://www.rana.lbl.gov/EisenSoftware.htm) (Eisen et al., 1998).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Forty-two embryos and four oocytes underwent a complete analysis of all nine genes. Five of the embryos were at the 23-cell stage, nine 45-cell, nine 68-cell, four 1011-cell, six were morulae and nine were blastocysts of which three had hatched. Additional embryos and oocytes were assessed for various combinations of these genes, but did not receive a full analysis. Combining data from partially and fully investigated samples, each gene was assessed in a minimum of 50 embryos and four oocytes. The most thoroughly investigated genes (BRCA1 and BRCA2) were analysed in 79 embryos and six oocytes.
Quantification of mRNA transcripts
The number of mRNA transcripts derived from each gene was ascertained by extracting RNA from each embryo, performing reverse transcription and subjecting the resultant cDNA to nine separate real-time PCR amplifications, one for each gene analysed. The average number of gene transcripts detected in embryos at different stages is given in Table I; an estimation of the number of transcripts per cell is also provided. Stage-specific data are presented in Figure 1. For most of the genes studied, expression levels were at their highest in oocytes and blastocysts. Remarkably, <50 transcripts per cell were found at several stages for a number of genes, indicating possible gene expression bottlenecks. For APC the 3145 transcripts detected in oocytes represented the greatest expression of this gene observed during this study. Similarly, the MAD2 gene also displayed higher expression levels in oocytes (averaging 3818 transcripts) than at other stages, with the exception of the most advanced (hatched) blastocysts.
|
|
All of the genes examined in this study displayed increased expression in 510-cell embryos relative to their 24-cell counterparts, presumably reflecting the onset of embryonic gene expression. In the cases of APC, BRCA2, BUB1, -actin and RB1 there was a 33.7-fold rise in average transcript number. For the ATM gene the rise was slightly higher (4.6-fold), but the greatest increase of all involved the BRCA1 gene. The average increase in expression for BRCA1 between the 24- and 510-cell stages was 12-fold. However, several embryos composed of 510 cells contained several thousand times more BRCA1 mRNA than those consisting of 24 cells. Interestingly, extreme BRCA1 expression was detected in approximately half the non-viable 510-cell embryos from St Barnabas Medical Center (n=15), but was not seen in any similarly staged embryos from the Mayo clinic (n=6). Large quantities of BRCA1 were also seen in some embryos at the 4-cell stage, the numbers presented in Table I representing an average. MAD2 and TP53 displayed a minor (1.5-fold) increase in transcript number between the 24- and 510-cell stages.
In 510-cell embryos the average number of transcripts per cell increased by 1.82.4-fold in the case of APC, BRCA2, BUB1, -actin, RB1 and ATM and by >9-fold in the case of BRCA1. However, the estimated number of transcripts from TP53 and MAD2 showed no increase per cell and in the case of MAD2 actually declined to 60% of the level detected in 24 cell embryos.
Progression from the 510-cell stage to morula formation was marked by stabilization in the expression of most of the genes. The APC, BRCA2, MAD2, BUB1, TP53, RB1 and ATM genes displayed small changes in average transcript number ranging from 0.9- to 1.9-fold. The only genes showing a large fluctuation during this period were the BRCA1 and -actin genes. BRCA1 transcripts underwent a 145-fold reduction in number, returning to more typical levels after having been so highly expressed in a subset of embryos at the 510-cell stage. The
-actin gene was the only gene to show a large increase in expression at the morula stage, with transcript numbers increasing 30-fold relative to embryos at the 510-cell stage.
The lack of a significant increase in mRNA during progression to the morula stage meant that the number of transcripts per cell actually fell for most of the genes. Average transcript numbers per cell dropped by 2070% for APC, BRCA2, MAD2, BUB1, TP53, RB1 and ATM. For BRCA1 the number of transcripts declined by a factor of 400, while -actin was the only gene to show a rise, with transcripts increasing almost 9-fold.
In contrast, blastocyst formation was accompanied by a large increase in the number of transcripts derived from each gene. This rise was most pronounced in fully expanded blastocysts that had emerged from the zona pellucida (hatched) (Figure 1). The average increases in transcript number exhibited by blastocysts ranged from 4-fold (BUB1) to 9-fold (RB1 and BRCA1). In most cases this translated into a small increase in the estimated number of transcripts per cell. The number of BUB1 transcripts per cell remained essentially unchanged in blastocysts compared with morulae, while the APC, BRCA2, MAD2, TP53, ATM and BRCA1 genes showed increases of between 1.2- and 1.9-fold. -actin and RB1 displayed greater increases of 2.4- and 3.5-fold respectively.
Identification of classes of embryos with similar gene expression profiles
A hierarchical cluster algorithm was employed in order to group together embryos displaying similar gene expression characteristics (Figure 2). The embryos analysed in this way were first coded, allowing the study to be performed in a blinded fashion. Only the 42 embryos and four oocytes that had each been tested for all nine genes were subjected to this analysis. Using this technique, five major groups showing clearly distinguishable gene expression profiles were identified. After decoding the samples to reveal their developmental stages, one group was found to comprise all embryos that had reached or passed the morula stage. This group was further split into two main branches, one encompassing the morulae and the other containing blastocysts. Blastocysts that had undergone hatching displayed a unique pattern of expression, very closely related to the expression of other blastocysts, yet distinct.
|
Association between the activities of individual genes
Hierarchical cluster analysis was also employed to examine whether each gene behaved in an independent fashion, or whether increases/decreases in the expression of one gene were accompanied by alterations in the expression of others. This analysis revealed that the genes investigated in this study could be placed into three main groups with related patterns of expression. The first group contained the ATM, RB1 and TP53 genes, the second included -actin, MAD2, BUB1 and APC and the third was made up of BRCA1 and BRCA2 (Figure 3).
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
We set out to characterize the expression of nine genes throughout human preimplantation development, using a highly accurate real-time RTPCR methodology. The suitability of this technique for the analysis of gene expression in individual oocytes and embryos has been previously established in our laboratory (Steuerwald et al., 1999). A number of investigations have examined the activity of single genes in individual oocytes or embryos (e.g. Pergament and Fiddler, 1998
and references therein; Hansis et al., 2001
; Taylor et al., 2001
; Spanos et al., 2002
; Patsoula et al., 2003
). However, very few have employed real-time PCR and to our knowledge no other study has accurately quantified the expression of so many genes within the same embryo. Because we were able to assess multiple genes in each sample, interplay between the activities of individual genes could be investigated. The nine genes included in this study were selected because they are known to have key roles in cell cycle regulation, DNA repair, signalling pathways, construction of the cytoskeleton and apoptosisall of which are of great importance during preimplantation development.
Variation in gene activity during preimplantation development
There were some similarities in expression shared by all of the genes examined. Transcripts from each of the nine genes were detectable at every stage, from mature oocyte to hatched blastocyst. However, the precise number of transcripts at different developmental stages varied considerably. The amount of mRNA derived for each gene was generally high in oocytes compared with cleavage stage embryos and morulae. Expression levels decreased dramatically after fertilization, recovered to a small extent between the 4- and 8-cell stages, underwent a further increase or in some cases a slight reduction at the morula stage, before increasing significantly at the blastocyst stage (Figure 1). In most cases the greatest gene expression was detected in hatched blastocysts. It is worth noting that the TP53, RB1 and ATM genes displayed considerably fewer transcripts in oocytes than other genes assessed, an interesting observation considering the key function of these genes in regulating apoptotic pathways and the cell cycle.
In terms of gene expression corrected for cell number, BUB1, APC, RB1, ATM and BRCA2 all showed a similar temporal pattern of gene activity. In these cases, low levels of expression per cell were observed in 23-cell embryos, morulae and blastocysts. For each gene a peak of expression, the greatest number of transcripts per cell seen post-fertilization, occurred at the 410-cell stage. BRCA1 also followed this pattern of expression, but displayed a peak of activity at the 410-cell stage that was much more pronounced than that of other genes.
The stage-specific pattern of expression displayed by TP53, MAD2 and -actin differed from the other genes examined. TP53 and MAD2 were expressed in a similar fashion, both displaying a more gradual decline in transcript number per cell following fertilization and a less obvious peak in expression at the 410-cell stage.
-actin also showed deviation from the typical pattern of preimplantation gene expression, displaying a small growth in transcript number at the 410-cell stage, rather than a clear peak in expression. Sharp increases in
-actin transcription occurred in morulae and became still more pronounced as embryos approached the blastocyst stage.
A significant reduction in mRNA transcript number follows fertilization
A significant decline in the mRNA content of oocytes following fertilization has been observed previously (Bachvarova and De Leon, 1980; Clegg and Piko, 1983
; Telford et al., 1990
). From our data it is apparent that transcripts from most genes reach extremely low levels in 2- and 3-cell embryos, in some cases scarcely above the threshold of detection. Reduction in transcript numbers between oocytes and 3-cell embryos ranged from 2- to 70-fold (averaging 17-fold) depending on the gene in question. This is consistent with the small amount of pre-existing data on human gene expression at this stage (Taylor et al., 2001
).
The gene that displayed the most significant decline in mRNA concentration following fertilization was APC (70-fold reduction in transcripts between oocytes and 23-cell embryos, leaving an average of just seven transcripts per cell). Whether this is due to targeted degradation of APC transcripts or a naturally short half-life of this mRNA species is unclear. APC expression begins to rise again after the 4-cell stage, but never regains the levels seen in oocytes. Along with BRCA1, APC was the only gene that had greater transcript numbers in oocytes than fully expanded, hatched blastocysts.
APC produces a large multifunctional protein that regulates the WNT signal transduction pathway, a pathway with important roles in early embryonic development and tumorigenesis (Munemitsu et al., 1995). The APC protein also binds to and stabilizes microtubules, regulates cytoskeletal function and may have a role in maintaining chromosome attachment to the mitotic spindle, through its interaction with BUB1 (Munemitsu et al., 1994
; Kaplan et al., 2001
). The high expression of APC in oocytes may indicate that this gene has particular significance during oocyte maturation or meiosis, perhaps stabilizing the attachment of chromosomes to spindle microtubules during meiosis II arrest. It will be interesting to determine whether the number of APC transcripts in oocytes declines as women age, potentially affecting oocyte and embryo aneuploidy rates, as has been suggested for other genes involved in ensuring accurate chromosome segregation (Steuerwald et al., 2000
).
The generalized depletion of mRNA during the first two cleavage divisions appears to be a normal occurrence and yet may have far-reaching consequences for the embryo. Many pathways essential for cellular health and homeostasis are controlled to a significant extent by the transcriptional activation or repression of specific genes. We speculate that, prior to activation of the embryonic genome, such pathways are inactive or exist in a more rigid form, controlled by a reservoir of proteins inherited from the oocyte. Thus, the embryo's ability to respond to environmental fluctuations during this period is likely to be limited. Additionally, suboptimal physiological conditions can lead to the degradation of specific proteins, perhaps leading them to fall below critical thresholds necessary for maintaining vital cellular functions. These factors combined have important implications for the practice of IVF and embryo culture.
In the current study the genes that displayed the steepest post-fertilization reduction in expression were TP53, RB1, ATM, BRCA2, BRCA1 and especially APC. Between the 2- and 4-cell stages the function of these genes may well be entirely dependent on previously synthesized maternally derived proteins. New synthesis of mRNA was only detected at, or in some cases shortly after, the 4-cell stage, consistent with previous data demonstrating the onset of embryonic gene activation (Tesarik et al., 1986; Braude et al., 1988
).
BRCA1 is expressed at high levels in a subset of 410-cell embryos
Some genes, such as BRCA1, displayed massive (several hundred-fold) increases in transcript number as early as the 4-cell stage. The BRCA1 protein is best known for its action in DNA repair pathways. However, it has other properties of potential relevance to early development, including a chromatin remodelling function (Ye et al., 2001) and the ability to associate with the XIST RNA responsible for X-chromosome inactivation in females (Ganesan et al., 2002
). The raised BRCA1 levels that we detected in 48-cell embryos may be of no significance to X-chromosome inactivation, as commitment to inactivation is not thought to occur until the embryo is composed of 1020 cells. However, it is interesting that approximately half of the 48-cell embryos from St Barnabas Medical Center displayed increased BRCA1, echoing the proportion of embryos expected to be female.
Fluctuations in gene expression between the 10-cell and blastocyst stages
Between the 10-cell and morula stages the expression of most genes stabilized or even underwent a small decline, relative to the 48-cell stage. Thereafter expression levels began to rise again, in relation to increasing cell number. Consequently, most genes displayed greater transcript numbers in blastocysts than at any other stage (although not necessarily a greater number of transcripts per cell). The high activity in blastocysts of the genes assessed in this study is not surprising. Most of these genes produce proteins that interact with DNA or chromosomes and consequently the quantity of protein required is likely to be closely related to the number of nuclei. Although the total volume of human preimplantation embryos changes little until after hatching from the zona pellucida, the number of cells and nuclei continue to increase steadily, with by far the largest number found in blastocysts.
RB1, BUB1 and -actin undergo significant up-regulation in hatched blastocysts
The increase in gene expression detected in hatched blastocysts was particularly pronounced for RB1, BUB1 and -actin, far exceeding any increases seen for these genes at other preimplantation stages. The RB1 gene product (pRB) plays an important role in regulating apoptosis and the cell cycle, performing its function through interaction with transcription factors, p53 and MDM2. RB1 is also thought to have a role during terminal differentiation in certain tissues (Zacksenhaus et al., 1996
). The increased expression of this gene in hatched blastocysts and the low levels seen at other preimplantation stages, particularly in oocytes, may be indicative of a role for RB1 in the increased apoptosis and/or differentiation of trophectoderm and inner cell mass seen during the latter stages of preimplantation development (Hardy, 1997
).
The high expression of BUB1 in hatched blastocysts compared to earlier stages may simply be the result of the great increase in cell number at this stage. The BUB1 protein binds to chromosomes during mitosis, as part of its role in the spindle assembly checkpoint (Jablonski et al., 1998; Ouyang et al., 1998). The larger number of cells in hatched blastocysts means that there are more chromosomes for BUB1 to interact with. Alternatively, the increased number of BUB1 transcripts may reflect rapid cell division occurring at this stage. Heightened BUB1 mRNA levels have been correlated with increased proliferation in some cell types (Pangilinan et al., 1997
; Ouyang et al., 1998
; Shigeishi et al., 2001).
One can speculate that the rapid increase in -actin expression seen at the morula and blastocyst stages is related to its role in the cytoskeleton. The reorganization of the cytoskeleton is important for compaction and the cytoskeleton has a vital function in initiating and maintaining cavitation in mammalian embryos (Capco and McGaughey, 1986
). In addition to the large scale changes in
-actin expression seen between embryos at different developmental stages, the number of transcripts also varied significantly between embryos of similar stage. This underlines that housekeeping genes cannot be relied upon to act as standards when quantifying gene expression in minute tissue samples. The levels of many housekeeping genes fluctuate considerably during progression through the cell cycle. Alterations in expression due to the phase of the cell cycle will be particularly apparent during developmental stages with relatively synchronous divisions, as multiple cells will simultaneously up-regulate or repress the same genes.
Embryos at specific developmental stages show characteristic patterns of gene expression
The collective data from embryos that had been assessed for all nine genes were also analysed using a hierarchical cluster algorithm, a mathematical method for identifying samples with similar characteristics. The results are displayed graphically as a dendrogram, embryos with similar gene expression being placed on closely associated branches of a tree diagram (Figure 2). Cluster analysis grouped embryos of specific stages together, confirming statistically that patterns of gene expression and developmental stage are closely associated. Similar clustering results were obtained using expression data adjusted to reflect cell number (i.e. number of transcripts per cell). This indicates that stage-specific clustering of samples was influenced by changes in the relative expression of individual genes and was not solely due to the generalized escalation in transcript numbers seen with advancing preimplantation development.
All embryos composed of <10 cells clustered together on a single major branch of the dendrogram. A sub-branch contained embryos presumed to be pre-genome activation (2-, 3- and some 4-cells). Morulae and blastocysts occupied separate but closely associated branches, while blastocysts that had undergone hatching were grouped together on a minor branch within the main blastocyst cluster. This suggested that it should be possible to define characteristic gene expression profiles for each stage of preimplantation development.
We performed a blind study to assess whether distinctive stage-specific patterns of gene expression could accurately estimate developmental stage (oocyte, 212-cell, morula, or blastocyst) and correctly categorized 37/46 embryos/oocytes (80%). Interestingly, more than half of the incorrectly classified embryos displayed a range of significant morphological abnormalities, suggesting that such aberrations may cause (or be caused by) perturbations of gene expression (Wells et al., 2005). It is likely that a more detailed analysis, looking at a larger number of genes, would provide an even more accurate assessment of stage of development.
Association between the expressions of individual genes
This study is unlike most other investigations of gene expression in human preimplantation embryos in that it provides data on multiple genes within the same embryo. This allowed a direct assessment of relationships between the activities of individual genes. Cluster analysis revealed genes that displayed coincident activation or repression, over and above the fluctuations attributable to developmental stage. The genes with the greatest parallel in expression were APC and BUB1, with MAD2 also showing similar activity.
MAD2 and BUB1 produce proteins that interact in the spindle assembly checkpoint, a cellular mechanism which acts to ensure accurate chromosome segregation (Li and Benezra, 1996; Ouyang et al., 1998
). Consequently, coincident expression of these genes is expected. Recent data from mice have demonstrated that the APC protein also forms a complex with the BUB1 product and facilitates the attachment of microtubules to kinetochores during mitosis (Kaplan et al., 2001
). The striking similarity in BUB1 and APC activity seen in this study suggests that the products of these genes also interact in human preimplantation embryos. Given the high incidence of chromosomal mosaicism seen in human embryos, the low expression levels of genes involved in maintaining accurate chromosome segregation during the 24 cell stage may be of significance (e.g. Delhanty et al., 1993
; Munné et al., 1993
; Harper et al., 1995
; Magli et al., 2000
; Wells and Delhanty, 2000
; Bielanska et al., 2002
).
Similarities in expression were also seen for the BRCA1 and BRCA2 genes. These genes produce proteins that have a role in maintaining genomic stability. Both have the ability to interact with numerous proteins and to form complexes involved in recognizing and subsequently repairing DNA (for review see Deng and Brodie, 2000). Specifically, it is thought that BRCA1 and BRCA2 proteins cooperate to control homologous recombination, an important mechanism for repairing double-stranded DNA breaks (Chen et al., 1998
).
Detection of high levels of BRCA1/BRCA2 expression may be indicative of embryos that contain significant quantities of damaged DNA. Raised expression of BRCA1 was particularly apparent in 48-cell embryos that were considered non-viable (derived from St Barnabas Medical Center), suggesting that DNA damage may be common in poor quality embryos at this stage. A larger study is currently underway to determine whether this is indeed the case. It has been suggested that excessive genetic damage is an important factor contributing to the failure of many embryos to progress beyond the cleavage stage of development. It is possible that DNA damage present at preimplantation stages originated in gametes (Henkel et al., 2004). DNA damage has been shown to be relatively common in sperm and may be correlated with poor implantation and pregnancy rates (Aravindan et al., 1997
; Larson-Cook et al., 2003
).
A further grouping of genes with similar expression, identified by cluster analysis, was ATM, RB1 and TP53. These genes produce multifunctional proteins that interact directly or indirectly in several essential pathways. Their functions include regulation of the cell cycle, directing pathways for DNA repair and apoptosis and controlling several cell-cycle checkpoints.
The only genes with an apparent inverse relationship regarding gene expression were TP53 and BRCA1. A previous study has demonstrated that BRCA1 expression is down-regulated by the presence of p53 (the TP53 protein product), indicating the existence of an intracellular p53/BRCA1 pathway in the response to a variety of conditions of stress (Arizti et al., 2000). Evidence suggests that BRCA1 protein modulates the action of p53, causing it to selectively activate DNA repair genes rather than apoptotic pathways. Thus, embryos expressing significant levels of TP53 but little BRCA1 may be undergoing apoptosis, while embryos with high levels of BRCA1 may be initiating DNA repair and/or cell cycle arrest.
Limitations of gene expression analysis
A change in the number of mRNA transcripts derived from a given gene is usually indicative of an alteration in the utilization of the protein that it produces. However, it is important to note that most if not all of the genes included in this study experience some degree of regulation at the post-translational level, through protein modification, degradation or sequestration. Thus, there may be occasions when a change in the concentration of active protein is not mirrored by an alteration in gene activity. Ultimately, findings based on the analysis of gene expression should be confirmed at the protein level.
![]() |
Conclusions |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The fluctuations in gene activity identified during this study provide an indication of some of the pathways active during specific phases of development. Further studies will undoubtedly result in a much improved understanding of the cellular processes occurring during the first few days of life. Greater knowledge in this area may aid the optimization of in vitro culture methods, as well as leading to new embryo viability assays. The recent successful analysis of single oocytes using microarray technology, which allows the simultaneous analysis of many thousands of genes, should accelerate the progress of discovery in this area (Bermudez et al., 2004).
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Aravindan GR, Bjordahl J, Jost LK and Evenson DP (1997) Susceptibility of human sperm to in situ DNA denaturation is strongly correlated with DNA strand breaks identified by single-cell electrophoresis. Exp Cell Res 236, 231237.[CrossRef][ISI][Medline]
Bachvarova R and De Leon V (1980) Polyadenylated RNA of mouse ova and loss of maternal RNA in early development. Dev Biol 74, 18.[CrossRef][ISI][Medline]
Bermudez MG, Wells D, Malter H, Munne S, Cohen J and Steuerwald NM (2004) Expression profiles of individual human oocytes using microarray technology. Reprod Biomed Online 8, 325337.[ISI][Medline]
Bielanska M, Tan SL and Ao A (2002) Chromosomal mosaicism throughout human preimplantation development in vitro: incidence, type, and relevance to embryo outcome. Hum Reprod 17, 413419.
Braude P, Bolton V and Moore S (1988) Human gene expression first occurs between the four- and eight-cell stages of preimplantation development. Nature 332, 459461.[CrossRef][ISI][Medline]
Capco DG and McGaughey RW (1986) Cytoskeletal reorganization during early mammalian development: analysis using embedment-free sections. Dev Biol 115, 446458.[CrossRef][ISI][Medline]
Chen J, Silver DP, Walpita D, Cantor SB, Gazdar AF, Tomlinson G, Couch FJ, Weber BL, Ashley T, Livingston DM and Scully R (1998) Stable interaction between the products of the BRCA1 and BRCA2 tumor suppressor genes in mitotic and meiotic cells. Molec Cell 2, 317328.[CrossRef][ISI][Medline]
Clegg KB and Piko L (1983) Poly(A) length, cytoplasmic adenylation and synthesis of poly(A)+ RNA in early mouse embryos. Dev Biol 95, 331341.[CrossRef][ISI][Medline]
Damario MA, Hammitt DG, Galanits TM, Session DR and Dumesic DA (1999) Pronuclear stage cryopreservation after intracytoplasmic sperm injection and conventional IVF: implications for timing of the freeze. Fertil Steril 72, 10491054.[CrossRef][ISI][Medline]
Delhanty JD, Griffin DK, Handyside AH, Harper J, Atkinson GH, Pieters MH and Winston RM (1993) Detection of aneuploidy and chromosomal mosaicism in human embryos during preimplantation sex determination by fluorescent in situ hybridisation, (FISH). Hum Mol Genet 2, 11831185.[Abstract]
Deng CX and Brodie SG (2000) Roles of BRCA1 and its interacting proteins. Bioessays 22, 728737.[CrossRef][ISI][Medline]
Edmonds DK, Lindsay KS, Miller JF, Williamson E and Wood PJ (1982) Early embryonic mortality in women. Fertil Steril 38, 447453.[ISI][Medline]
Eisen MB, Spellman PT, Brown PO and Botstein D (1998) Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95, 1486314868.
Eldadah ZA, Grifo JA and Dietz HC (1995) Marfan syndrome as a paradigm for transcript-targeted preimplantation diagnosis of heterozygous mutations. Nat Med 1, 798803.[CrossRef][ISI][Medline]
Ganesan S, Silver DP, Greenberg RA, Avni D, Drapkin R, Miron A, Mok SC, Randrianarison V, Brodie S, Salstrom J et al. (2002) BRCA1 supports XIST RNA concentration on the inactive x chromosome. Cell 111, 393405.[CrossRef][ISI][Medline]
Gardner DK, Lane M, Stevens J and Schoolcraft WB (2001) Noninvasive assessment of human embryo nutrient consumption as a measure of developmental potential. Fertil Steril 76, 11751180.[CrossRef][ISI][Medline]
Gianaroli L, Magli MC, Ferraretti AP, Fiorentino A, Garrisi J and Munné S (1997) Preimplantation genetic diagnosis increases the implantation rate in human in vitro fertilization by avoiding the transfer of chromosomally abnormal embryos. Fertil Steril 68, 11281131.[CrossRef][ISI][Medline]
Hansis C, Tang YX, Grifo JA and Krey LC (2001) Analysis of Oct-4 expression and ploidy in individual human blastomeres. Mol Hum Reprod 7, 155161.
Hardy K (1997) Cell death in the mammalian blastocyst. Mol Hum Reprod 3, 919925.[Abstract]
Harper JC, Coonen E, Handyside AH, Winston RM, Hopman AH and Delhanty JD (1995) Mosaicism of autosomes and sex chromosomes in morphologically normal, monospermic preimplantation human embryos. Prenat Diagn 15, 4149.[ISI][Medline]
Henkel R, Hajimohammad M, Stalf T, Hoogendijk C, Mehnert C, Menkveld R, Gips H, Schill WB and Kruger TF (2004) Influence of deoxyribonucleic acid damage on fertilization and pregnancy. Fertil Steril 81, 965972.[CrossRef][ISI][Medline]
Kaplan KB, Burds AA, Swedlow JR, Bekir SS, Sorger PK and Nathke IS (2001) A role for the Adenomatous Polyposis Coli protein in chromosome segregation. Nat Cell Biol 3, 429432.[CrossRef][ISI][Medline]
Larson-Cook KL, Brannian JD, Hansen KA, Kasperson KM, Aamold ET and Evenson DP (2003) Relationship between the outcomes of assisted reproductive techniques and sperm DNA fragmentation as measured by the sperm chromatin structure assay. Fertil Steril 80, 895902.[CrossRef][ISI][Medline]
Li Y and Benezra R (1996) Identification of human mitotic checkpoint gene: hsMAD2. Science 274, 246248.
Magli MC, Jones GM, Gras L, Gianaroli L, Korman I and Trounson AO (2000) Chromosome mosaicism in day 3 aneuploid embryos that develop to morphologically normal blastocysts in vitro. Hum Reprod 15, 17811786.
Munemitsu S, Souza B, Muller O, Albert I, Rubinfeld B and Polakis P (1994) The APC gene product associates with microtubules in vivo and promotes their assembly in vitro. Cancer Res 54, 36763681.[Abstract]
Munemitsu S, Albert I, Souza B, Rubinfeld B and Polakis P (1995) Regulation of intracellular beta-catenin levels by the adenomatous polyposis coli (APC) tumor-suppressor protein. Proc Natl Acad Sci USA 92, 30463050.
Munné S, Lee A, Rosenwaks Z, Grifo J and Cohen J (1993) Diagnosis of major chromosome aneuploidies in human preimplantation embryos. Hum Reprod 8, 21852191.[Abstract]
Munné S, Sandalinas M, Escudero T, Velilla E, Walmsley R, Sadowy S, Cohen J and Sable D (2003) Improved implantation after preimplantation genetic diagnosis of aneuploidy. Reprod Biomed Online 7, 9197.[Medline]
Ouyang B, Lan Z, Meadows J, Pan H, Fukasawa K, Li W and Dai W (1998) Human Bub1: a putative spindle checkpoint kinase closely linked to cell proliferation. Cell Growth Diff 9, 877885.[Abstract]
Pangilinan F, Li Q, Weaver T, Lewis BC, Dang CV and Spencer F (1997) Mammalian BUB1 protein kinases: map positions and in vivo expression. Genomics 46, 379388.[CrossRef][ISI][Medline]
Patsoula E, Loutradis D, Drakakis P, Michalas L, Bletsa R and Michalas S (2003) Messenger RNA expression for the follicle-stimulating hormone receptor and luteinizing hormone receptor in human oocytes and preimplantation-stage embryos. Fertil Steril 79, 11871193.[CrossRef][ISI][Medline]
Pergament E and Fiddler M (1998) The expression of genes in human preimplantation embryos. Prenat Diagn 18, 13661373.[CrossRef][ISI][Medline]
Spanos S, Rice S, Karagiannis P, Taylor D, Becker DL, Winston RM and Hardy K (2002) Caspase activity and expression of cell death genes during development of human preimplantation embryos. Reproduction 124, 353363.
Steuerwald N, Cohen J, Herrera RJ and Brenner CA (1999) Analysis of gene expression in single oocytes and embryos by real-time rapid cycle fluorescence monitored RTPCR. Mol Hum Reprod 5, 10341039.
Steuerwald N, Cohen J, Herrera RJ and Brenner CA (2000) Quantification of mRNA in single oocytes and embryos by real-time rapid cycle fluorescence monitored RTPCR. Mol Hum Reprod 6, 448453.
Taylor DM, Handyside AH, Ray PF, Dibb NJ, Winston RM and Ao A (2001) Quantitative measurement of transcript levels throughout human preimplantation development: analysis of hypoxanthine phosphoribosyl transferase. Mol Hum Reprod 7, 147154.
Telford NA, Watson AJ and Schultz GA (1990) Transition from maternal to embryonic control in early mammalian development: a comparison of several species. Mol Reprod Dev 26, 90100.[CrossRef][ISI][Medline]
Tesarik J, Kopecny V, Plachot M and Mandelbaum J (1986) Activation of nucleolar and extranucleolar RNA synthesis and changes in the ribosomal content of human embryos developing in vitro. J Reprod Fertil 78, 463470.[ISI][Medline]
Verlinsky Y, Cieslak J, Freidine M, Ivakhnenko V, Wolf G, Kovalinskaya L, White M, Lifchez A, Kaplan B, Moise J et al. (1996) Polar body diagnosis of common aneuploidies by FISH. J Assist Reprod Genet 13, 157162.[ISI][Medline]
Wells D and Delhanty JD (2000) Comprehensive chromosomal analysis of human preimplantation embryos using whole genome amplification and single cell comparative genomic hybridization. Mol Hum Reprod 6, 10551062.
Wells D, Bermudez MG, Steuerwald N, Malter HE, Thornhill AR and Cohen J (2005) Association of abnormal morphology and altered gene expression in human preimplantation embryos. Fertil Steril (in press).
Wittwer CT, Herrmann MG, Moss AA and Rasmussen RP (1997) Continuous fluorescence monitoring of rapid cycle DNA amplification. BioTechniques 22, 130138.[ISI][Medline]
Ye Q, Hu YF, Zhong H, Nye AC, Belmont AS and Li R (2001) BRCA1-induced large-scale chromatin unfolding and allele-specific effects of cancer-predisposing mutations. J Cell Biol 155, 911921.
Zacksenhaus E, Jiang Z, Chung D, Marth JD, Phillips RA and Gallie BL (1996) pRb controls proliferation, differentiation, and death of skeletal muscle cells and other lineages during embryogenesis. Genes Dev 10, 30513064.[Abstract]
Submitted on September 7, 2004; accepted on January 12, 2005.