Silencing CENPF in bovine preimplantation embryo induces arrest at 8-cell stage

Conserved molecular portraits of bovine and human blastocystsas a consequence of the transition from maternal to embryoniccontrol of gene expression

James Adjaye,1 Ralf Herwig,1 Thore C. Brink,1 Doris Herrmann,2 Boris Greber,1 Smita Sudheer,1

Detlef Groth,1 Joseph W. Carnwath,2 Hans Lehrach,1 and Heiner Niemann2

1Department of Vertebrate Genomics, Max Planck Institute for Molecular Genetics, Berlin; and 2Department ofBiotechnology, Institute for Animal Science, Neustadt, Germany

Submitted 23 February 2007; accepted in final form 25 June 2007

Adjaye J, Herwig R, Brink TC, Herrmann D, Greber B,Sudheer S, Groth D, Carnwath JW, Lehrach H, Niemann H.Conserved molecular portraits of bovine and human blastocysts as aconsequence of the transition from maternal to embryonic control ofgene expression. Physiol Genomics 31: 315–327, 2007. First pub-lished June 26, 2007; doi:10.1152/physiolgenomics.00041.2007.—The present study investigated mRNA expression profiles of bovineoocytes and blastocysts by using a cross-species hybridization ap-proach employing an array consisting of 15,529 human cDNAs asprobe, thus enabling the identification of conserved genes duringhuman and bovine preimplantation development. Our analysis re-vealed 419 genes that were expressed in both oocytes and blastocysts.The expression of 1,324 genes was detected exclusively in theblastocyst, in contrast to 164 in the oocyte including a significantnumber of novel genes. Genes indicative for transcriptional andtranslational control (ELAVL4, TACC3) were overexpressed in theoocyte, whereas cellular trafficking (SLC2A14, SLC1A3), proteasome(PSMA1, PSMB3), cell cycle (BUB3, CCNE1, GSPT1), and proteinmodification and turnover (TNK1, UBE3A) genes were found to beoverexpressed in blastocysts. Transcripts implicated in chromatinremodeling were found in both oocytes (NASP, SMARCA2) andblastocysts (H2AFY, HDAC7A). The trophectodermal markers PSG2and KRT18 were enriched 5- and 50-fold in the blastocyst. Pathwayanalysis revealed differential expression of genes involved in 107distinct signaling and metabolic pathways. For example, phosphati-dylinositol signaling and gluconeogenesis were prominent pathwaysidentified in the blastocyst. Expression patterns in bovine and humanblastocysts were to a large extent identical. This analysis comparedthe transcriptomes of bovine oocytes and blastocysts and provides asolid foundation for future studies on the first major differentiationevents in blastocysts and identification of a set of markers indicativefor regular mammalian development.

oocytes; microarrays; cross-species hybridization

CRITICAL STEPS during early development such as the timing offirst cell division, activation of the embryonic genome,compaction, blastocyst formation, expansion, and hatchingare regulated by a well-orchestrated expression of genes.During the transition from maternal to embryonic control ofdevelopment, maternal transcripts (common to both theunfertilized oocyte and the preimplantation embryo) are

depleted and embryo-specific transcripts involved in orches-trating early embryogenesis are generated in a processknown as maternal to embryo transition (MET). Proteinsynthesis is required for MET, which occurs at the 2-cellstage in mouse and between the 4- and 8-cell stages inhuman and the 8- and 16-cell stages in bovine embryos (12,72, 73). Consistent with the need to synthesize appropriatetranscription factors and regulatory proteins (34, 43, 68, 69),the poly(A) content per mouse embryo increases fivefoldfrom the two-cell stage to the blastocyst stage (17). To date,the expression pattern of only 200 –250 of the estimated25,000 genes of the mammalian genome have been studiedby reverse transcriptase-polymerase chain reaction (RT-PCR) in bovine embryos (72, 73). Advances in cDNA arraytechnology have made possible large-scale analysis of thetranscriptome of the mammalian genome in a quantitativemanner. However, integration of suitable genomic methodsand bioinformatic tools is a prerequisite for this kind ofanalysis (1, 33).

Bovine oocytes and blastocysts contain �2.4 ng and �5ng of total RNA, respectively (9). Therefore, it is necessaryto pool large amounts of biological material and/or amplifythe RNA without significantly altering the relative levels ofindividual mRNAs. Protocols for effective and nonbiasedamplification of RNA from single oocytes and embryos havebeen developed (1, 2, 4, 11), and transcript profiling ofmurine, porcine, and bovine oocytes and embryos has beenreported (21, 27, 30, 52, 58, 64, 70). We previously dem-onstrated the feasibility and high reproducibility of cross-species analysis using a human cDNA microarray. Messen-ger RNAs derived from human and bovine fetal brains werecompared, and the correlation coefficient of cross-hybrid-ization between orthologous genes was 0.94 (3). We nowreport the first comparative transcriptome profile of bovineoocytes and blastocysts using a human 15,529 gene chip(the ENSEMBL chip) as an initial step to gain insight intogene expression and regulation during preimplantation de-velopment. These stages are representative of maternal andembryo-controlled expression, and specifically the blasto-cyst represents the first major differentiation event in devel-opment (1, 2, 4, 11). We have identified numerous genesexpressed in the oocyte and blastocyst and have revealedtheir associated signaling and metabolic pathways. Theseresults provide a basis for functional studies related to theimportance of these genes and related pathways in preim-plantation development.

Article published online before print. See web site for date of publication(http://physiolgenomics.physiology.org).

Address for reprint requests and other correspondence: J. Adjaye, MaxPlanck Institute for Molecular Genetics (Dept. of Vertebrate Genomics),Ihnestrasse 73, 14195 Berlin, Germany (e-mail: [email protected]); H.Niemann, Institute for Animal Science, (Dept. of Biotechnology), Mariensee,31535 Neustadt, Germany (e-mail: [email protected]).

Physiol Genomics 31: 315–327, 2007.First published June 26, 2007; doi:10.1152/physiolgenomics.00041.2007.

1094-8341/07 $8.00 Copyright © 2007 the American Physiological Society 315

MATERIALS AND METHODS

Clone selection and microarray fabrication. The selection of the15,529 cDNA clones and fabrication of the arrays was as previouslydescribed (4).

In vitro production of bovine embryos. Bovine embryos wereproduced as described recently with minor modifications (74). Briefly,ovaries from a local slaughterhouse were transported to the laboratoryin Dulbecco’s phosphate-buffered saline (PBS; no. D6650, Sigma, St.Louis, MO) at 25–30°C. Cumulus-oocyte complexes (COC) wereisolated via slicing (29). Category I COC (with a homogeneous evenlygranulated cytoplasm possessing at least 3 layers of compact cumuluscells) and category II COC (with �3 layers of cumulus cells orpartially denuded but also with a homogeneous evenly granulatedcytoplasm) were pooled in TCM-air [TCM 199 containing L-glu-tamine and 25 mM HEPES (Sigma) supplemented with 22 �g/mlpyruvate, 350 �g/ml NaHCO3, 50 �g/ml gentamicin, and 0.1%bovine serum albumin (BSA; fraction V, no. A9647, Sigma)].

For maturation in vitro, TCM 199 containing L-glutamine and 25mM HEPES served as basic medium. One milliliter was supplementedwith 22 �g of pyruvate, 2.2 �g of NaHCO3, and 50 �g of gentamicin.For oocyte maturation, this medium was supplemented with 10 IU ofequine chorionic gonadotropin and 5 IU of human chorionic gonad-otropin (Suigonan; Intervet, Tonisvorst, Germany) and 10% estruscow serum (collected at day 1 of standing estrus). COC were dividedinto groups of 20 –25, transferred into 100-�l maturation dropsunder silicone oil, and cultivated in a humidified atmospherecomposed of 5% CO2 in air at 39°C for 24 h. Maturation wasdetermined by extrusion of the first polar body and expansion ofthe cumulus cells (29).

After in vitro maturation, COC were rinsed in fertilization medium(Fert-TALP supplemented with 6 mg/ml BSA) and fertilized inFert-TALP containing 10 �M hypotaurine (Sigma), 1 �M epinephrine(Sigma), and 0.1 IU/ml heparin (Serva) and 6 mg/ml BSA. Frozensemen from one bull with proven fertility in in vitro fertilization (IVF)was used. The final sperm concentration added per fertilization dropwas 1 � 106 sperm/ml. Fertilization was initiated during a 19-hcoincubation under the same temperature and gas conditions asdescribed for maturation.

Presumptive zygotes were cultured in 30 �l of synthetic oviductfluid medium supplemented with 6% BSA after complete removal ofthe adhering cumulus cells by repeated pipetting. Embryos werecultured in vitro in a mixture of 7% O2, 88% N2, and 5% CO2 (AirProducts, Hattingen, Germany) in Modular incubator chambers (no.615300, ICN Biomedicals, Aurora, OH). Blastocysts were harvestedat day 7 of development (day 0 � IVF). After being washed threetimes in PBS containing 0.1% polyvinyl alcohol, oocytes and blasto-cysts were stored in pools of 10 for oocytes or pools of 3 forblastocysts at �80°C in a minimum volume (�5 �l) of medium untilRNA extraction.

Sampling strategy and messenger RNA extraction. Pooled oocyteand blastocyst samples were made up of 40 in vitro-matured oocytesand 12 in vitro-derived blastocysts, respectively. By including bio-logical material from four different in vitro production runs (4 � 10oocytes, 4 � 3 blastocysts) we could cover the biological variabilityand concomitantly obtain sufficient amounts of RNA (�60 and 90 ngRNA, respectively) from both developmental stages for the amplifi-cation. Template mRNA from pooled samples was obtained with aDynabead oligo(dT) extraction kit (Dynal Biotech). The procedureincluded the following steps. Twenty microliters of lysis/bindingbuffer was added, and the samples were processed according to themanufacturer’s protocol with the following modification: beads werewashed once with washing Buffer A and three times with washingBuffer B before the elution of mRNA was performed at 65°C in 11 �lof RNase-free water.

First-strand synthesis. One microliter of 20 �M T7 poly(dT)21

primer (5�-TCTAGTCGACGGCCAGTGAATTGTAATACGACTC-

ACTATAGGGCGTTTTTTTTTTTTTTTTTTTTT-3�) was added tothe eluted template mRNA, and the samples were denatured by a2-min incubation at 70°C. Samples were immediately placed on icefor 1 min, and then 8 �l of master mix was added, consisting of 2 �lof 10� RT buffer (Applied Biosystems, Foster City, CA), 2 �l of 50mM MgCl2 (Invitrogen), 2 �l of 10 mM dNTP solution (AmershamBiosciences), 1 �l (20 U) of RNasin (Applied Biosystems), and 1 �l(50 U) of Moloney murine leukemia virus (MMLV) reverse transcrip-tase (Applied Biosystems), followed by incubation at 42°C for 1 h.

Second-strand synthesis. For second-strand synthesis, 50 �l of 2�DOP-PCR buffer (Roche Diagnostics, Penzberg, Germany), 1 �l of 40�M degenerate oligonucleotide primer (5�-CCGACTCGAGNNNN-NNATGTGG-3�; Roche Diagnostics), and 23.5 �l of water were addedto each 20-�l reverse transcriptase sample. Samples were heated to 95°Cfor 5 min and then incubated at 30°C for 2 min. After the addition of 5IU Taq DNA polymerase (Invitrogen), the samples were heated at a rateof 0.2°C/s to 72°C and incubated at 72°C for 3 min.

After second-strand synthesis was completed, 2.5 �l each of T7poly(dT)21 primers and degenerate oligonucleotide primers wereadded. Twenty-five PCR cycles were performed as follows: cDNAwas denatured at 94°C for 30 s, and primer annealing was performedat 60°C for 30 s, followed by an elongation step at 72°C for 4 min.

Purification and concentration. PCR products were purified withthe GFX-PCR DNA purification kit (Amersham Biosciences) accord-ing to the manufacturer’s protocol and eluted in 75 �l of 0.1�Tris-EDTA (TE) buffer, and volumes were adjusted to 12 �l byvacuum centrifugation.

In vitro transcription. Two microliters of buffer was added to thesolution after centrifugation, followed by 2 �l each of ATP, CTP,GTP, UTP, DTT, and T7 RNA polymerase (Ampliscribe/Biozym).In vitro transcription was carried out at 42°C for 3 h, followed byDNase I digestion at 37°C for 15 min. Amplified RNA was purifiedwith the RNeasy Mini kit (Qiagen, Hilden, Germany).

This procedure provides a means of amplifying minute amounts ofRNA from single embryos with minimal bias; the correlation coeffi-cient is 0.97 for autonomous amplifications of RNA from differentembryos (11).

Direct labeling of RNA and hybridizations. MIAME (MinimumInformation About Microarray Experiments) guidelines were adheredto in our experimental design.

Four independent labeling (including dye swaps, Cy3 and Cy5)reactions and hybridizations per amplified RNA (aRNA) sample werecarried out, always using 3 �g of aRNA for each biological sample.This corresponds to four technical replicates for oocytes and blasto-cysts, respectively. Full details of labeling and hybridization reactions,slide washing (0.2� SSC-0.1% SDS, 55°C), and scanning are de-scribed elsewhere (3, 4).

Global data analysis. Data was normalized in two steps. The firststep accounted for the dye effects caused by the difference in red/green fluorescence labeling for each single experiment. Here we usedthe LOWESS method (18). The parameters d, W, t, and f are neededto adjust the procedure. The order of the polynomial d and the numberof iterations t were set to 1. The weighting function W was set asrecommended; the fitting parameter f was set to 0.3. In the second stepof the normalization each single-array experiment was normalized toa common median value, taking into account additional multiplicativenoise. The procedure has been described previously (4).

The validity of gene expression of each individual signal wasjudged by comparison to a negative control sample. To verify whethera given gene was significantly expressed, we compared its signal to asignal distribution derived from negative controls. In our array design,we distributed �3,362 empty spot positions on the array. Afterquantification, a small, nonzero intensity was assigned to each emptyspot reflecting the amount of background signal on the array. Sincethese positions were spread uniformly over the array, the distributionof signals reflects a global background distribution for the experimentand indicates whether cDNA signals were at or above the background

316 TRANSCRIPTOMIC ANALYSIS OF BOVINE EARLY DEVELOPMENT

Physiol Genomics • VOL 31 • www.physiolgenomics.org

level of expression. For each cDNA, we counted the relative propor-tion of empty positions on the array that were smaller than theobserved intensity [background (BG) tag]. BG tags from replicatedexperiments for the same cDNA were averaged. Thus high values(close to 1) indicated that the cDNA was expressed in the tissue tested,whereas low values reflected noise. cDNAs were considered “ex-pressed” when their average BG tag was above 0.9, a thresholdconsistent with the limit of visual detection of the spots.

For each cDNA, we performed statistical tests based on the repli-cate signals in each oocyte and blastocyst hybridized. Three standardtests were used in parallel: Student’s t-test, the Welch test, andWilcoxon’s rank-sum test (37). For each stage, i.e., oocyte vs. blas-tocyst, we performed at least four independent replicate hybridiza-tions.

Pathway analysis. Array data were used to test whether entire groupsof genes associated with specific pathways showed differential expres-sion. Pathways were taken from the KEGG database (version KEGG2,01/06/2005). The procedure has been described previously (4).

Independent verification of array data. A description of the primersequences, amplicon length, and annealing temperatures is given inSupplemental Table 1.1

End-point RT-PCR. Conventional end-point PCR was performedwith 25 ng of the cDNA for each gene or 50 fg of cDNA of the globincontrol. The 50-�l reaction mixture consisted of 5 �l of 10� PCRbuffer (Invitrogen), 1.5 mM MgCl2, each dNTP at 200 �M, and eachprimer at 0.5 �M. Amplification was performed in an MJ ResearchPTC-200. Hot start was accomplished by adding 1 U of Taq DNApolymerase (Invitrogen) while the reaction mixture was maintained at72°C. The PCR program consisted of denaturization at 97°C for 2min; 2 min at 72°C, at which point Taq polymerase was added; andthen cycles consisting of 95°C for denaturization, annealing at atemperature specific to each primer pair for 15 s, and elongation at72°C for 15 s. The number of cycles performed was determined foreach gene in preliminary experiments to give measurements that camewithin the linear phase of the amplification curve. PCR products wereseparated on a 2% agarose gel in Tris-borate-EDTA (TBE) buffercontaining 0.2 �g/ml ethidium bromide.

Real-time RT-PCR. Four biological replicates each consisting ofpools of two oocytes or blastocysts were thawed in 40 �l of lysisbuffer. Two picograms of rabbit globin mRNA (BRL, Gaithersburg,MD) was added to the solution to serve as an internal standard.Poly(A)� RNA was isolated with a Dynabeads mRNA Direct Kit(Dynal, Oslo, Norway) and was eluted with 22 �l of distilled (d)H2O.Eleven microliters (equivalent to a single oocyte or blastocyst) wasused as input for the reverse transcription reaction, and the other 11 �lwas used as input for the negative control (to monitor genomic DNAcontamination, without reverse transcriptase).

Reverse transcription was performed in a 20-�l volume consistingof 2 �l of 10� RT buffer (Invitrogen), 2 �l of 50 mM MgCl2(Invitrogen), 2 �l of 10 mM dNTP solution (Amersham Biosciences),1 �l (20 U) of RNasin (Applied Biosystems), 1 �l (50 U) of MMLVreverse transcriptase (Applied Biosystems), and 1 �l of hexamers (50�M) (Applied Biosystems). The samples were incubated at 25°C for10 min for primer annealing and then incubated at 42°C for 1 h.Finally, the samples were heated to 95°C for 5 min. Two-microliteraliquots of the RT reaction were used as template for real-time PCR,which was performed in 96-well optical reaction plates (AppliedBiosystems). The PCR mix in each well included 10 �l of 2� PowerSYBR Green PCR Master Mix (Applied Biosystems), 6.4 �l of dH2O,0.8 �l each of the forward and reverse primers (5 �M), and 2 �l (0.1oocyte/blastocyst equivalent) of cDNA in a final reaction volume of20 �l. Rabbit globin (100 fg) was amplified along with the targetgenes for normalization. Each gene was analyzed four times.

The reaction was carried out in an ABI 7500 Fast Real-TimeSystem (Applied Biosystems) using the following program: denatur-ation and activation of the Taq polymerase for 10 min at 95°C,followed by 40 cycles of 95°C for 15 s and 60°C for 1 min, and finallyheating with a ramp rate of 2% at 95°C for 15 s, 60°C for 15 s, and95°C for 15 s to display a dissociation curve of the product. Datagenerated by Sequence Detection Software 1.3.1 were transferred toMicrosoft Excel for analysis. Differential mRNA expression of eachgene was calculated by the comparative threshold cycle (Ct) methodrecommended by the manufacturer.

Functional annotations. For functional annotations, genes specifi-cally expressed in bovine oocytes and blastocysts were identified bysubtracting the overlapping subset from the individual lists of signif-icantly expressed genes. The corresponding lists of Entrez gene IDswere then employed as input for the DAVID gene annotation tool(25). Several GO categories specifying the biological processes (GO:0008150), molecular functions (GO: 0003674), and cellular compo-nent (GO: 0005575) were selected, and relative abundances of geneswithin these categories were computed to compare the oocyte- andblastocyst-specific gene expression profiles.

Online database. To enable a global overview of gene expressionin oocytes and blastocysts, which can be interrogated, we have presentedthe expression data as a database for searching for expression levelsof specific genes and their related Gene Ontologies (http://goblet.molgen.mpg.de/cgi-bin/stemcell/preimplantation-development.cgi). We used GO terminology taken from the Gene Ontologywebsite (http://www.geneontology.org). This was imported intothe sqlite database (http://www.sqlite.org). Data analysis wascarried out with R-Statistics software (http://www.r-project.org).

RESULTS

Global data analysis. Amplified RNA from the oocyte pool(40 oocytes) and the blastocyst pool (12 blastocysts) were usedfor expression profiling. Before carrying out the global expres-sion analysis, we tested the generated amplified oocyte andblastocyst RNA samples by end-point RT-PCR for preserva-tion of known oocyte and blastocyst markers (Fig. 1). Thepurpose of the end-point PCR was to ensure that our amplifi-cation approach had not adversely skewed the relative levels ofoocyte and blastocyst marker genes. This was necessary beforethe enormous task of microarray analysis. The expression ofoocyte markers ZAR1 and GDF9 (49, 75) and blastocystmarkers CSNK1A1 and PITX2 (4) were retained. We thenproceeded with the large-scale expression analysis. Four tech-nical replicate hybridizations were performed for each RNA,including Cy dye swaps. The reproducibility of hybridization

1 The online version of this article contains supplemental material.Fig. 1. End-point RT-PCR analysis of known oocyte- and blastocyst-specificgenes to test for the fidelity of mRNA isolation and amplification.

317TRANSCRIPTOMIC ANALYSIS OF BOVINE EARLY DEVELOPMENT


experiments between replicates was assessed by calculatingPearson’s correlation coefficients. These values range from0.83 to 0.98 for identical cell types, thus indicating a highdegree of reproducibility in array hybridization (data notshown). Data were normalized as described in MATERIALS AND

METHODS (Fig. 2A). To judge whether a given gene is expressedin these cell types, we computed a BG tag for its signal. Thisnumber reflects the proportion of background noise in relationto the actual signal (3). Typically, a BG tag of 0.9 indicates adetectable signal for the probe (Fig. 2B). Using this criterion,we found that 1,959 (12%) of genes represented as probes onthe chip were detected in both cell types. As shown in Fig. 2B,most of these genes were expressed in the blastocyst (1,324) orwere common to both cell types (419). The number of genesdetected as expressed exclusively in the oocyte was rather low[166 (12%)]. This unexpectedly low number of oocyte-expressed genes detected might be due partly to the fact thatthe comparative sequence homology for the majority of thegenes presented as probes on the array is low and thus unde-tected as expressed with the high stringency of the hybridiza-tion protocol previously described (3). In support of thisfinding is the fact that in our previous studies using the samearray platform (4) we detected �6,000 genes expressed in thehuman blastocyst, whereas only 1,324 were detected as ex-pressed in the bovine blastocyst.

The full data set comprising all results is presented as adatabase for interrogating the expression levels of specific genesand their related Gene Ontologies (http://goblet.molgen.mpg.de/cgi-bin/stemcell/preimplantation-development.cgi).

Distinct and overlapping gene expression in oocytes andblastocysts. Despite the fundamental importance of the transi-tion from maternal to embryo-coded gene transcription,progress in understanding how human and bovine preimplan-tation embryonic cells establish regulation of transcription andtheir own pattern of gene expression from a completely inac-tive genome has been hampered by the paucity of moleculesknown to regulate these processes. To gain further insights intothe molecular mechanisms underpinning this process, we ana-lyzed our data set for putative genes essential for oogenesis(maternal transcripts), the transition from maternal to embryo

control of gene expression (MET), and further growth of theembryo to the blastocyst stage (embryonic transcripts). Toenable identification of these transcripts, three distinct tests(Student’s t-test, Welch test, Wilcoxon’s rank-sum test) wereused to help overcome individual bias (37). This analysisrevealed a subset of 1,220 putative marker genes that weredifferentially expressed at the 0.05 level of significance. A fulllist of oocyte and blastocyst markers and genes common toboth cell types is presented in Supplemental Tables 2, 3, and 4,respectively. The expression patterns of a selection of theseoocyte and blastocyst marker genes were verified indepen-dently on replicate (4�) unamplified mRNA samples of oocyteand blastocyst origin. Of these genes, only COPS4 showeddiscordant expression patterns between the array-derived dataand real time. The results shown in Fig. 3 confirmed thatindeed DAZL and UPF1 were enriched in oocytes, whereasVDAC2 and EI24 were enriched in blastocysts. A descriptionof the primer sequences and amplicon lengths are given inSupplemental Table 1.

Fig. 2. Global data analysis. A: effect of LOWESS normal-ization. Plotted are the ratio of the red (R) and green (G)signals for each spot (log scale, y-axis) and the signal range(log scale, x-axis) of a typical experiment. Whereas the rawdata show a nonlinear bias in particular at the extremes of thesignal area (left), after LOWESS normalization this bias iseliminated (right). B: Venn diagram of genes expressed inthe different cell types. Gene expression was judged by anumerical value [background (BG) tag] computed from anegative control sample (right).

Fig. 3. Real-time PCR confirmation of the array data. Ratios of gene expres-sion in blastocysts (BL) and oocytes (OOC) are plotted for real-time PCR andmicroarray data. Log2 ratios are plotted on the y-axis with positive denotingoverexpression in oocytes (DAZL, COPS4, and UPF1) and negative denotingoverexpression in blastocysts (EI24 and VDAC2).



Conserved gene expression patterns in human and bovineblastocysts. To demonstrate this, we used these marker genesto perform hierarchical cluster analysis in order to identifycoexpressed groups of genes in bovine oocytes and blastocysts.Included in this analysis was the recently published data set ongenes expressed in human blastocysts (4). The analysis re-vealed greater similarities between the transcriptomes of hu-man and bovine blastocysts, in contrast to bovine oocytes (Fig.4A), and likewise we identified sets of genes enriched in bovineoocytes compared with both human and bovine blastocysts(Fig. 4B).

Maternal transcription. Of the 297 putative oocyte markergenes, 34 (11.45%) were functionally uncharacterized andtherefore could be designated as novel. These novel genesconsisted of several zinc finger-containing transcription fac-tors. A short list of the 20 most abundantly expressed tran-scripts in the oocyte compared with the blastocyst is given inTable 1. The expression profiles of several translational regu-lators and RNA binding factors were identified. For example,ELAVL4 was found to be 24-fold enriched. This gene belongsto a family of RNA binding proteins with crucial roles intranslational regulation of stored maternal mRNAs duringoocyte-to-embryo transition (30). Another gene, TACC3, witha similar role, was 14-fold overexpressed, contrasting with theexpression of TACC2, which was barely detectable although ithad been shown to be expressed in mouse oocytes (30). Thisdiscrepancy might likely be due to the limited detection sen-sitivity of microarray platforms compared with RT-PCR.Nonetheless, TACC1 was not detected as expressed in bothbovine and mouse oocytes. Interestingly, genes previouslyannotated as testis germ line-specific transcripts were also

enriched in oocytes; these include DAZL, TTTY12, TEX9,YBX2 (MYS2), and PHTF1. Of these, TTTY12, TEX9, andPHTF1 are of unknown function.

Blastocyst transcription. Of the 916 putative blastocystmarker genes identified, 137 (14.9%) are uncharacterized andtherefore can be designated as novel; among these were severalzinc finger-containing putative transcription factors. The 20most abundantly expressed transcripts in the blastocyst com-pared with the oocyte are listed in Table 2.

A feature of postfertilization transcription is high levels ofTATA-less promoter activity as exemplified by Eif1A, which ishighly transcribed during MET in the mouse (23). The probefor the human ortholog EIF1A was not represented on thearray; nonetheless, several members of the eukaryotic transla-tion initiation factor gene family (EIF2, EIF3, EIF4, and EIF5)were overexpressed in the blastocyst. For example, EIF2B2was 33-fold enriched, while mouse Eif2 had been shown to beexpressed at low levels in metaphase II (MII) oocytes and tohave increased expression in blastocysts (40). Members of theTEA-DNA binding domain transcription factors (HTEAD genefamily), TEAD1 and TEAD2, were barely detectable, reflectingtheir low-level expression, which had been demonstrated in themouse by RT-PCR (41). However, HTEAD3 was 2.5-foldenriched. Genes encoding the poly(A) binding protein inter-acting protein 2 (PAIP2), the poly(A) binding protein C1(PABPC1), which has well-known roles in mRNA translationand decay in the cytoplasm, and also PABPN1, which associ-ates with nuclear pre-mRNAs, were enriched in the blastocyst.A whole range of transcripts encoding ribosomal proteins(RPLs) and subunits (RPSs) such as RPL5 and RPS19 were 25-and 20-fold enriched. Also present were transcripts encoding

Fig. 4. Clusters of genes that show overexpression in oocytes (A) and human and bovine blastocysts (B). Colors correspond to normalized signals. For each gene,signals were divided by the average gene signal across all conditions (log scale). Red boxes indicate that the signal in the particular condition is higher than theaverage signal, whereas green boxes indicate the opposite. Hierarchical clustering was performed on 1,220 marker genes that showed differential expression,using Pearson correlation as a pairwise similarity measure and average linkage as an update rule. The analysis was done with J-Express Pro 2.6 Software(www.molmine.com).



the RNA binding motif proteins (RBM25, RBM16, RBMX).These proteins contain arginine-proline-arginine and RNA-recognition motifs present in proteins involved in regulationof nuclear pre-mRNA splicing (10). As anticipated, genesinvolved in posttranslation modifications such as the kinases(TNK1, TYRO3), UDP-glucose ceramide glucosyltransferase(UGCG), ubiquitinases (UBE3A, UBB, UBEA1, UBE2V2,UBA2), transport (XPO1/EXPORTIN), and proteasomes(PSMA1, 2, 3, 6; PSMB3, 5, 7, 8; and PSMC2, 5, 6 subunits)were overexpressed in blastocysts.

Mitochondria are the most abundant organelles in the mam-malian oocyte and early embryo and play a key role in anumber of physiological events during development, includingthe very first stages following fertilization (65). We detectedexpression of several transcripts encoding mitochondrial ribo-somal protein subunits (MRPL2, 17, 18, 27, 28, 39, 48 andMRPS17). This would suggest active translation of mitochon-

drial proteins. However, the significance of the overrepresen-tation of the L subunit is unknown. We also found markers forregulators of intracellular pH and maintenance of NAD�/NADH balance and an enrichment of the NADH dehydroge-nases (NDUFAB1; NDUFB3, 8, 9, 11; NDUFS2, 4, 8).

Several members of cell cycle-related genes (CCNAR,CCND, CCNE1, CCNF, CCNG, CCNJ, CCNK, CDC16,CDC37, CDK9, CHEK1, CHES1, CLK2, CLK4, GSPT1) wereoverexpressed in the blastocyst. In mammalian cells, cyclin E(CCNE1)-CDK2 complexes are activated in the late G1 phaseof the cell cycle and are believed to play an essential role inpromoting S-phase entry (51, 66). In addition, GSPT1, aGTP-binding protein essential for the G1-to-S phase transitionof the cell cycle, is also upregulated (38). Eighteen genes thatencode proteins that function as solute carriers were enriched inthe blastocyst. This large family of solute carriers facilitatestransport of glucose (SLC2A14), nucleotides (SLC25A4 and

Table 1. Twenty most abundant transcripts in oocyte

Gene Description Accession No. Enrichment

SLC25A12 Solute carrier family 25 (mitochondrial carrier, Aralar), member 12 AA477641, AA479540 1.8E�04RUNX1T1 Runt-related transcription factor 1; translocated to, 1 (cyclin D-related) H46432, H46979 1.4E�04MAPK8IP3 Mitogen-activated protein kinase 8 interacting protein 3 AA504284 1.2E�04HOXA3 Homeobox A3 AI796896 8.8E�03Novel Transcribed locus AI217693 8.2E�03COPS4 COP9 constitutive photomorphogenic homolog subunit 4 (Arabidopsis) AA134596, AA134597 3.4E�03AUTS2 Autism susceptibility candidate 2 W61304, W65368 1.3E�03LOC401494 Similar to RIKEN 4933428I03 AI804932 9.3E�02ARHGAP25 Rho GTPase activating protein 25 H60815, H60902 9.0E�02C10orf45 Chromosome 10 open reading frame 45 N67028, W04157 4.8E�02PXMP3 Peroxisomal membrane protein 3, 35 kDa (Zellweger syndrome) W86121, W86122 3.2E�02SIX1 Sine oculis homeobox homolog 1 (Drosophila) N79004 3.0E�02PCDHGC3 Protocadherin subfamily C, 3 R34362 2.9E�02SIPA1L2 Signal-induced proliferation-associated 1-like 2 AA035016, AA035484 2.7E�02HAS2 Hyaluronan synthase 2 AI142961 2.2E�02GPR19 G protein-coupled receptor 19 H07878, H07970 2.0E�02MGC14156 Hypothetical protein MGC14156 AA042945, AA043063 1.7E�02DAZL Deleted in Azoospermial-Like AA774538 1.5E�02CNTNAP2 Contactin associated protein-like 2 R51354, R51460 1.5E�02RENT1 Regulator of nonsense transcripts 1 H09150, H09207 1.5E�02

Table 2. Twenty most abundant transcripts expressed in the blastocyst

Gene Description Accession No. Enrichment

ALOX12P2 Arachidonate 12-lipoxygenase pseudogene 2 AA411432, AA411550 5.3E�02VDAC2 Voltage-dependent anion channel 2 T66813 5.0E�02ARHGAP15 Rho GTPase activating protein 15 NM018460 3.2E�02EI24 Etoposide induced 2.4 mRNA AI016141 2.8E�02C1orf37 Chromosome 1 open reading frame 37 R01008, R01009 2.5E�02EI24 Etoposide induced 2.4 mRNA AI016141 1.3E�02TPT1 Tumor protein, translationally controlled 1 AI735372 1.2E�02SFRS10 Splicing factor, arginine/serine-rich 10 (transformer 2 homolog, Drosophila) N26496, N35548 1.2E�02EDNRB Endothelin receptor type B H28710, H28840 1.1E�02WEE1 WEE1 homolog (Schizosaccharomyces pombe) H44948, H44994 1.0E�02MGC23918 Coiled-coil domain containing 12 AA497045 1.0E�02KIAA1238 KIAA1238 protein AI611175 9.9E�01LOC203427 Similar to solute carrier family 25 , member 16 DN999924 9.1E�01OR51E1 Olfactory receptor, family 51, subfamily E, member 1 NM152430 8.7E�01Novel Transcribed locus AI215550 8.5E�01RPL23AP7 Ribosomal protein L23a pseudogene 7 BC000596, BE275192 8.5E�01RPS3A Predicted: Homo sapiens similar to ribosomal protein S3a; 40S ribosomal protein S3a NM001006 8.4E�01KIAA0423 KIAA0423 N92639, W39148 8.3E�01DKFZP586M1120 Leucine rich repeat containing 48 AI183684 7.8E�01GNB2L1 Guanine nucleotide binding protein (G protein), polypeptide 2-like 1 NM006098 7.8E�01



SLC25A5), thiamine (SLC19A2), cations (SLC7A7), zinc(SLC39A7), and glutamate (SLC1A3). Finally, the trophectodermmarkers PSG2 and KRT18 identified in our previous study (4)were 5- and 50-fold enriched, respectively.

Genes common to both oocyte and blastocyst. Of the 419commonly expressed genes, 11 (2.6%) are uncharacterized andtherefore can be designated as novel. A list of the 20 mostabundantly expressed transcripts common to both the blasto-cyst and the oocyte is presented in Table 3. During mammalianpreimplantation development, the gametic epigenetic informa-tion is reprogrammed in a process that is essential for theestablishment of nuclear totipotency in oocytes and two-, four-,and eight-cell stage embryos, for directing appropriate expres-sion of imprinted genes, and also for establishing differentia-tion to two distinct cell layers in the blastocyst, the inner cellmass and the trophoblast (4). We looked specifically for theexpression of genes implicated in this process (Fig. 5). Thegene SMARCA2/BRM, which is a member of the SWI/SNFfamily highly similar to the brahma protein of Drosophila, was89-fold enriched in the oocyte. The encoded protein is part ofthe large ATP-dependent chromatin remodeling complex SNF/SWI, which regulates both transcriptional activation and re-pression (7). Other epigenetic regulators such as the methyl-transferases SET7 and DNMT1 (oocyte form DNMT1o) wereenriched in the blastocyst and the oocyte, respectively. Severalcore histones (HIST1H2BD), histone acetylases (MYST4), anddeacetylases (HDAC7A) were differentially expressed in theoocytes and blastocysts, respectively (Fig. 5A). In keeping withepigenetic modifications and the effect on gene transcription,we also screened the data set for differential expression ofputatively imprinted genes (Fig. 5B). The data suggest signif-icant overexpression of MEST in oocytes, while GNAS,UBE3A, SNRPN, and PON2 were enriched in the blastocyst.These patterns of expression have also been observed in bothhuman and bovine preimplantation embryos (5, 55, 57).

Functional annotation of genes expressed in oocyte andblastocyst. Functional classification of gene expression profilesaccording to their biological process would give significantinsights into changes in molecular function, biological pro-

cesses, and associated cellular components during the transi-tion from maternal to embryonic control of transcription duringpreimplantation development. With the DAVID gene annota-tion tool (25, 36), several GO categories specifying the bio-logical processes (GO: 0008150) in which the gene productsare involved as well as their molecular functions (GO:0003674) and cellular component (GO: 0005575) were se-lected and relative abundances of genes within these categoriescomputed to compare the oocyte- and blastocyst-specific geneexpression profiles (Fig. 6).

For example, regarding molecular function, oxidoreductaseactivity and structural molecule activity were overrepresentedin the blastocyst. Organic acid metabolism and protein biosyn-thesis are the most overrepresented biological processes in theblastocyst. Regarding cellular component, genes of the outermembrane are overrepresented in oocytes. The genes ex-pressed in oocytes and blastocysts together with their respec-tive Gene Ontology terms are available in our database (http://goblet.molgen.mpg.de/cgi-bin/stemcell/preimplantation-development.cgi).

Oocyte enriched genes conserved in human, bovine, andmouse. To identify oocyte-conserved genes we made use ofavailable data sets that compared mouse oocyte transcripts tothose of humans (44) and compared these to our oocyte-expressed genes (Supplemental Tables 2 and 4). This resultedin the identification of 31 genes; among these was BUB3,which has been shown in Xenopus egg extracts to be essentialfor the mitotic spindle checkpoint pathway (14). Other com-mon transcripts were the germ cell-specific transcripts DAZLand NASP and a panel of genes implicated in cell proliferationsuch as FER, KIT, and MAPRE1. The full list and descriptionof these genes are given in Supplemental Table 5.

Signaling and metabolic pathways operative in oocytes andblastocysts. To delineate metabolic and signaling pathwaysfundamental for understanding basic mechanisms regulatingearly development, we searched the oocyte and blastocystexpression data for differential expression of components ofpathways by assigning P values with Wilcoxon’s matched-pairsigned-rank test, as described in MATERIALS AND METHODS. This

Table 3. Twenty most abundant transcripts expressed in both oocytes and blastocysts

Gene Description Accession No. Ratio

PAXIP1L PAX interacting (with transcription-activation domain) protein 1 N21080, N27916 0.94COPS7B COP9 constitutive photomorphogenic homolog subunit 7B (Arabidopsis) AA451853, AA452053 0.96CCT5 Chaperonin containing TCP1, subunit 5 (�) R73042, R73043 0.96LOC150l678 Myeloma overexpressed 2 AI891108 0.97LGALS4 Lectin, galactoside-binding, soluble, 4 (galectin 4) BC005146 0.98ARHGEF2 Rho/rac guanine nucleotide exchange factor (GEF) 2 NM004723 0.98DCTN6 Dynactin 6 H78562, H79048 0.99CX3CR1 Chemokine (C-X3-C motif) receptor 1 R34908, R49299 0.99FLJ11011 Hypothetical protein FLJ11011 AK001873 0.99MR-1 Myofibrillogenesis regulator 1 R06739, R06788 0.99LIAS Lipoic acid synthetase AJ224162, N24865, N31804 1.00KIAA0931 PH domain and leucine-rich repeat protein phosphatase-like AA099359, AA101908 1.00MGC14595 Hypothetical protein MGC14595 AA056028, AA056046 1.01FUSIP1 FUS interacting protein (serine-arginine rich) 1 H11042, H11130 1.01AURKB Aurora kinase B R97911, R97912 1.01KLF3 Kruppel-like factor 3 (basic) R06749, R06798 1.02RNF7 Ring finger protein 7 N63421, N95727 1.03NIF3L1BP1 Ngg1 interacting factor 3 like 1 binding protein 1 R65638, R67018 1.04RHEB Ras homolog enriched in brain T99524, T99583 1.04C11orf23 Chromosome 11 open reading frame 23 H64859, H65457 1.05



method is distinct from the commonly employed strategies foridentifying differentially expressed genes by repeated measure-ments with a two-sample location test, such as Student’s t-testor Wilcoxon’s rank-sum test (4). The analysis shows, forexample, that oxidative phosphorylation, cell cycle, apoptosis,

focal adhesion, calcium, integrin, WNT, MAPK, transforminggrowth factor (TGF)-, NOTCH, and phosphatidylinositolsignaling pathways and metabolic processes such as glycolysis,sterol biosynthesis, and androgen and estrogen metabolism areoperative during preimplantation development. The full list of

Fig. 5. Expression of genes involved in epigenetic regula-tion and chromatin remodeling. Genes involved in chroma-tin remodeling (A) and imprinted genes (B) are shown.Array-derived ratios of differential gene expression in blas-tocyst and oocytes are plotted. Log2 ratios are plotted on they-axis, with positive values denoting overexpression inoocytes and negative values denoting overexpression inblastocyst. *Significant change, with a P value �0.05.

Fig. 6. Functional annotation of bovine oocyte and blastocyst-specific genes based on cross-species microarray hybridizations. Human orthologous gene IDs ofgenes specifically expressed in bovine oocytes and blastocysts were fed into the DAVID gene annotation tool (25, 36) and classified according to molecularfunction (left), biological process in which the gene products are involved (center), and cellular component (right). Bars represent relative abundances of geneswithin selected GO categories.



107 signaling and metabolic pathways identified by theseanalyses is provided in Supplemental Table 6. Significantlyaltered pathways as judged by differential expression of constit-uent genes in the oocyte and blastocyst are shown in Table 4.These signaling and metabolic pathways had been already

shown to be operative in the human blastocyst (4), thusconfirming the conservation of these developmentally impor-tant pathways in mammalian preimplantation development.Furthermore, ATP production via the oxidative phosphoryla-tion pathway has been shown to be essential for bovine embryo

Table 4. Signaling and metabolic pathways operative in oocytes and blastocysts

Pathway No. of Genes Z Score P Value Oocyte Up Blastocyst Up

Ribosome 47 5.84 2.6E�09 3 44Oxidative phosphorylation 45 5.03 2.5E�07 6 39Focal adhesion 28 3.12 9.1E�04 7 21Calcium signaling pathway 23 3.01 1.3E�03 4 19Pyrimidine metabolism 13 2.83 2.3E�03 1 12Phosphatidylinositol signaling system 11 2.67 3.8E�03 1 10Glycerophospholipid metabolism 12 2.59 4.8E�03 2 10Protein export 8 2.52 5.9E�03 0 8Regulation of actin cytoskeleton 27 2.43 7.6E�03 7 20Purine metabolism 26 2.40 8.2E�03 5 21Citrate cycle (TCA cycle) 8 2.38 8.6E�03 1 7Toll-like receptor signaling pathway 7 2.37 9.0E�03 0 7Inositol phosphate metabolism 8 2.24 1.3E�02 1 7Lysine degradation 11 2.22 1.3E�02 3 8Prostaglandin and leukotriene metabolism 6 2.20 1.4E�02 0 6Tryptophan metabolism 18 2.16 1.6E�02 6 12Complement and coagulation cascades 8 2.10 1.8E�02 7 1Tyrosine metabolism 10 2.09 1.8E�02 2 8Biosynthesis of steroids 5 2.02 2.2E�02 0 5Fatty acid metabolism 15 1.99 2.3E�02 4 11-Hexachlorocyclohexane degradation 10 1.99 2.3E�02 2 8Glycolysis/gluconeogenesis 12 1.88 3.0E�02 3 9ATP synthesis 7 1.86 3.1E�02 2 5Proteasome 18 1.85 3.2E�02 3 15Tight junction 25 1.84 3.3E�02 6 19Glutathione metabolism 4 1.83 3.4E�02 0 4Chondroitin/heparan sulfate biosynthesis 4 1.83 3.4E�02 0 4Amyotrophic lateral sclerosis (ALS) 6 1.78 3.7E�02 1 5Bile acid biosynthesis 5 1.75 4.0E�02 1 4Glutamate metabolism 5 1.75 4.0E�02 1 4Cell cycle 27 1.68 4.6E�02 9 18Wnt signaling pathway 22 1.64 5.1E�02 7 15Fatty acid biosynthesis (path 2) 3 1.60 5.4E�02 0 3Porphyrin and chlorophyll metabolism 3 1.60 5.4E�02 0 3Terpenoid biosynthesis 3 1.60 5.4E�02 0 3RNA polymerase 3 1.60 5.4E�02 0 3Dentatorubropallidoluysian atrophy (DRPLA) 3 1.60 5.4E�02 0 3MAPK signaling pathway 47 1.59 5.6E�02 14 33Pyruvate metabolism 8 1.54 6.2E�02 1 7Valine, leucine, and isoleucine degradation 7 1.52 6.4E�02 2 5Pentose and glucuronate interconversions 5 1.48 6.9E�02 1 4Neurodegenerative disorders 9 1.48 6.9E�02 2 7Ubiquitin-mediated proteolysis 9 1.48 6.9E�02 3 6-Alanine metabolism 4 1.46 7.2E�02 1 3Propanoate metabolism 4 1.46 7.2E�02 1 3Alanine and aspartate metabolism 2 1.34 9.0E�02 0 2Selenoamino acid metabolism 2 1.34 9.0E�02 2 0Styrene degradation 2 1.34 9.0E�02 0 2Methane metabolism 2 1.34 9.0E�02 0 2Reductive carboxylate cycle (CO2 fixation) 2 1.34 9.0E�02 0 2Riboflavin metabolism 2 1.34 9.0E�02 0 2Alkaloid biosynthesis I 2 1.34 9.0E�02 0 2Alkaloid biosynthesis II 2 1.34 9.0E�02 2 0

Array data were used to test whether entire groups of genes associated with specific pathways show differential expression between oocytes and blastocysts.Details of the pathways were taken from the KEGG database (version KEGG2, 01/06/2005). The procedure has been described previously (4). Gene expressionwas compared in the blastocyst and oocyte stage and the differences of array signals were used for computing Wilcoxon’s paired signed-rank test. Genes thatwere judged as nondetectable by the background (BG) value criterion were excluded from analysis. Pathway, pathway description; no. of genes, number of genesthat were taken into account for computing the statistical test; Z score, standard normal approximation of the test statistic; P value, P value of the standard normaldistribution for the respective Z score; oocyte up, number of genes that have higher expression in oocytes than in blastocysts; blastocyst up, number of genesthat have higher expression in blastocysts than in oocytes. All pathways are shown, but it should be noted that the normal approximation is valid for samplesizes �25.



development in vitro (62). Integrin and calcium signaling, onthe other hand, have been shown to be required for efficientblastocyst implantation in the mouse (67), and the requirementof the MAPK signaling pathways has been shown to beessential for mouse preimplantation (67).

DISCUSSION

We previously demonstrated (3) the feasibility and repro-ducibility of cross-species hybridization using bovine RNA aslabeled targets on a human cDNA array as a means of identi-fying evolutionarily conserved genes and pathways. Here,using the same approach but on an array consisting of 15,529human cDNAs, we have identified conserved developmentallyregulated genes and associated signaling and metabolic path-ways operative in unfertilized human and bovine oocytes andin vitro-derived blastocysts. The qualitative and quantitativenature of the data set were evaluated with three standard testsin parallel: Student’s t-test, the Welch test, and Wilcoxon’srank-sum test.

The ultimate approach for identifying genes solely tran-scribed by the newly formed embryonic genome and not thefemale gamete is to block protein synthesis by culturing one- totwo-cell embryos in -amanitin. This approach has revealedembryonic transcription of Hsp70.1 (16), U2afbp-rs (46),eIF-4C (22), Xist (77), and Tead2 (42) and from the transcrip-tion-requiring complex (TRC) (20) in the mouse. While theseearlier studies were informative, the advent of new analyticalmethods has enabled the identification of an increasing numberof genes as being characteristic of either maternal or embryonicexpression. Among the methods used to study human, bovine,and porcine preimplantation development are subtractive hy-bridization, expressed sequence tag (EST) libraries, differentialdisplay, serial analysis of gene expression (SAGE), in situ datamining, and microarrays (1).

Although each technique has its own unique merits, DNAmicroarrays have gained precedence because they enable com-parative whole genome transcriptome analysis, albeit at alower sensitivity of detection. A major drawback with the useof microarrays compared with the other technologies for thesort of analysis we describe here is that low-abundance oocyte-and preimplantation embryo-specific transcripts may not berepresented as probes on most microarray platforms. Addition-ally, deadenylated mRNAs, which are known to be stored asinactive transcripts in oocytes, would never be isolated ortranscribed with oligo(dT)-mediated mRNA labeling proto-cols, and therefore these genes within the oocyte pool willnever be presented as labeled targets for possible detection onthe arrays. This drawback is clearly manifested in our analysis,where we detected only 583 genes expressed in the oocytecompared with 1,743 in the blastocyst (Fig. 2B). Nonetheless,we identified conservation of 31 of these genes in human,bovine, and mouse oocytes. This unexpectedly low number canbe attributed to the fact that the mouse and human oocyte datasets were derived from the Affymetrix platform and were notbased on a cross-species approach.

The cDNA array used in this study has an advantage forcross-species comparison in that the length of each cDNAprobe is sufficiently long that a small degree of mismatch istolerated. In the present study, high stringency was used, whichimplies that transcripts with low homology between the bovine

and human sequences are not detected as expressed. On theother hand, a strong signal in either the oocyte or the blastocysthybridization demonstrates high nucleotide sequence homol-ogy, which in turn indicates conservation of an important genecomponent in an essential signaling or metabolic pathway.Another major advantage of our cross-species hybridization ap-proach is that developmentally conserved novel and annotatedmarker genes could be identified for human and bovine oocytesand blastocysts. In addition, we have assigned each gene to itscorresponding GO molecular function, biological process, andcellular component (http://goblet.molgen.mpg.de/cgi-bin/stemcell/preimplantation-development.cgi). The final func-tional classifications of the observed gene profiles according tobiological process will give significant insight into the molec-ular changes involved in compaction and cell adhesion occur-ring during the transition from maternal to embryonic controlof transcription.

In this study, we have demonstrated the identification ofgenes, related Gene Ontologies, and pathways relevant in theoocyte and the blastocyst. Gene expression patterns in theoocyte, with the overexpression of ELAVL4, TACC3, andMSY2, confirmed the importance of transcriptional and trans-lation regulation (30, 76). Interestingly, the expression ofELAV4 and TACC3 mirrors observations in mouse oocytes (30,35). Furthermore, TACC3 and TACC2 are functional homologsof MASKIN and, like the ELAVL gene family, are involved intranslational control in oocytes and preimplantation embryos.Because translational control plays a central role during oocytematuration and early embryogenesis, it is not surprising thatthese genes show evolutionary conservation.

High-level expression of genes previously identified as germcell-specific transcripts (DAZL, TTTY12, TEX9, YBX2/MSY2,PHTF1) in oocytes suggests that a developmentally conservedprocess is operative in both male and female germ cell devel-opment (56, 63). Although testes and ovary are functionallynonequivalent, they share a common meiosis machinery, andthese transcripts probably play similar roles in male and femalegerm cells. Lack of DAZL expression leads to embryonic arrestof germ cell development in mice (48), and DAZL is expressedin human oocytes (44, 56). Sequence variants of DAZL andsingle nucleotide polymorphisms are associated with prema-ture ovarian failure and menopause (63). The gene encodingMSY2 is a member of a multifunctional Y-box protein familyimplicated in regulating the stability and translation of mater-nal mRNAs during mouse oogenesis. Mouse Msy2 transcriptand protein are expressed in growing oocytes and one-cellembryos, but subsequently are degraded by the late two-cellstage, with no detectable expression in the blastocysts. More-over, results from RNA interference-mediated suppression ofMsy2 function in mouse oocytes support its role in stabiliz-ing maternal mRNAs in growing oocytes, a process essentialfor generating meiotically and developmentally competentoocytes (76).

Genes indicative of transcriptional, translational control, andposttranslational modifications were found to be active in theblastocyst. This is manifested by the 33-fold enrichment oftranscripts for EIF2B2 and the 2.5-fold enrichment of tran-scripts for HTEAD3. This implies conservation of the functionof these gene families in mammalian preimplantation develop-ment (40, 41). Also enriched in the blastocysts is XPO1/EXPORTIN, which is involved in signal-mediated transport of



proteins from the nucleus and has been shown to be enrichedin mouse and swine embryos at the two-cell and four-cellstages, respectively (32). Furthermore, it has been speculatedthat XPO1 may regulate MET, possibly by controlling the poolof transcription factors present in the nucleus (13). High levelsof expression of numerous components of the proteasome,specifically the PSMA, PSMB, and PSMC class of proteins,lends further support to the notion that RNA and proteindegradation are an integral part of the regulatory machineryessential for MET, as already shown in the mouse (30).Together, these expression patterns underscore the importanceof pre-mRNA processing, stability, mRNA trafficking, andtranslation seen in other mammalian cell types (39).

Epigenetic regulation and chromatin remodeling as an un-derlying mechanism of transcriptional control necessary forexecuting MET (59) have been aptly illustrated by the differ-ential expression of key genes known to be involved in thismode of transcriptional regulation (Fig. 5), for example, theoverexpression in the oocyte of SMARCA2/BRM and the oo-cyte variant of DNMT1, together with overexpression of themethyltransferase (SET7), histone acetylases (MYST4), anddeacetylases (HDAC7A). Furthermore, this would indicate thatthe establishment and maintenance of MET involves a certaindegree of methylation of CpG islands, histone methylation,acetylation, and deacetylation resulting in transcriptional pro-grams essential for executing MET and subsequent develop-ment to the blastocyst stage. These epigenetic modificationswould presumably also account for the observed differentialexpression of a subset of putatively imprinted genes (Fig. 5B).It is also probable that the proposed epigenetic changes areacting in trans to alter the expression of these putativelyimprinted genes. In conclusion, the observed expression pat-terns reflect the essential roles played by these genes in theregulation of X inactivation, imprinting, maintenance of plu-ripotency, and establishment of the trophectodermal lineages inpreimplantation embryos (4, 31).

The identified signaling and metabolic pathways operative inbovine oocytes and blastocysts mirror observations in humanblastocysts (Table 4; Ref. 4). For example, the observedincrease in glucose metabolism in the blastocyst of both spe-cies is supportive of existing data showing that during the earlystages of mammalian preimplantation development embryosare characterized by low levels of respiration and glucosemetabolism (8, 26) and glucose may even compromise devel-opment of early human embryos (19). After compaction, pre-implantation embryos are capable of metabolizing glucose, andthis switch seems to be associated with developmental regula-tion of the expression of the glucose transporters and theirisoforms during preimplantation development (6, 15), which isin line with the observed increased expression of genes encod-ing glucose transporters such as SLC2A14 in the blastocyst.Furthermore, numerous studies in bovine, human, and mousesuggest that phosphatidylinositol 3-kinase signaling regulatesglucose utilization in blastocysts and therefore is essential fortheir survival (4, 6, 54). In contrast, the carrier SLC25A12, aCa2�-binding mitochondrial aspartate-glutamate carrier, is en-riched in the oocyte. Although the significance of this ispresently unknown, it could be an indication of the requirementof intracellular Ca2� as a universal signal to trigger metabolicactivity in oocytes (28). Since the function of most of the novelfamily members is unknown, future research must concentrate

on defining their function, e.g., substrate specificity, regulationof subcellular localization, and mechanisms of cellular traffick-ing during preimplantation development.

Bovine preimplantation development has recently gainedsignificant attention because of its similarity with human de-velopment. Bovine embryos are now easily produced in largenumbers from slaughterhouse material and serve as a valuablemodel in assisted reproductive technologies (ARTs) includingin vitro embryo production and somatic cloning, which are inan advanced stage in this species (45). Numerous studies haveshown that the use of ARTs may be associated with a varietyof pathological symptoms that are summarized under the term“large offspring syndrome” (47, 71). Epidemiologic studies inhumans have revealed that the use of some reproductivetechnologies is associated with an increased frequency ofimprinting defects, twins, and neurological disorders (24, 53,60, 61). The causative mechanism is thought to be deviantepigenetic control of mRNA expression patterns during pre-implantation development that persist throughout fetal devel-opment (50). The present study was performed with pooled invitro-matured oocytes and in vitro-cultured blastocysts derivedfrom IVF of in vitro-matured oocytes. This unavoidable andcommonly used strategy of pooling oocytes and preimplanta-tion-stage embryos for microarray-based gene expression stud-ies does not reveal transcriptional heterogeneity between thesesamples as a result of culturing in vitro. Furthermore, some ofthe mRNA expression profiles shown here will need verifica-tion with in vivo-matured oocytes and in vivo-produced blas-tocysts. While such studies cannot be performed with humanembryos, the bovine model is a practical alternative. We haveinitiated such studies by hybridizing RNA from all relevantpreimplantation-stage embryos derived in vivo to the bovineAffymetrix array.

In conclusion, our cross-species approach has demonstratedthat the transition from maternal to embryonic control oftranscription in mammals (bovine and human) is accomplishedby the regulated expression of developmentally conservedgenes and related Gene Ontologies. This process is influencedand regulated by changes in the oocyte and blastocyst epig-enomes, leading to altered chromatin architecture and ulti-mately induced or repressed gene transcription and associatedsignaling and metabolic pathways necessary for preimplanta-tion development.

ACKNOWLEDGMENTS

We are grateful to Dr. Marie-Laure Yaspo and the German Resource Centrefor Genome Research (RZPD, Berlin) for providing the cDNA clones and toDr. Claus Hultschig for printing the slides.

GRANTS

This work was supported by the Max Planck Society, the German NationalGenome Research Network, (NGFN-01GR0105), and Deutsche Forschungs-gemeinschaft (Ni 256/27-1 and AD 184/5-1).

REFERENCES

1. Adjaye J. Whole-genome approaches for large-scale gene identificationand expression analysis in mammalian preimplantation embryos. ReprodFertil Dev 17: 37–45, 2005.

2. Adjaye J, Bolton V, Monk M. Developmental expression of specificgenes detected in high-quality cDNA libraries from single human preim-plantation embryos. Gene 237: 373–383, 1999.

3. Adjaye J, Herwig R, Herrmann D, Wruck W, Benkahla A, Brink TC,Nowak M, Carnwath JW, Hultschig C, Niemann H, Lehrach H.



Cross-species hybridisation of human and bovine orthologous genes onhigh density cDNA microarrays. BMC Genomics 5: 83, 2004.

4. Adjaye J, Huntriss J, Herwig R, BenKahla A, Brink TC, Wierling C,Hultschig C, Groth D, Yaspo ML, Picton HM, Gosden RG, LehrachH. Primary differentiation in the human blastocyst: comparative molecularportraits of inner cell mass and trophectoderm cells. Stem Cells 23:1514–1525, 2005.

5. Assou S, Anahory T, Pantesco V, Le Carrour T, Pellestor F, Klein B,Reyftmann L, Dechaud H, De Vos J, Hamamah S. The human cumulus-oocyte complex gene-expression profile. Hum Reprod 21: 1705–1719,2006.

6. Augustin R, Pocar P, Navarrete-Santos A, Wrenzycki C, Gandolfi F,Niemann H, Fischer B. Glucose transporter expression is developmen-tally regulated in in vitro derived bovine preimplantation embryos. MolReprod Dev 60: 370–376, 2001.

7. Banine F, Bartlett C, Gunawardena R, Muchardt C, Yaniv M, Knud-sen ES, Weissman BE, Sherman LS. SWI/SNF chromatin-remodelingfactors induce changes in DNA methylation to promote transcriptionalactivation. Cancer Res 65: 3542–3547, 2005.

8. Barbehenn EK, Wales RG, Lowry OH. The explanation for the block-ade of glycolysis in early mouse embryos. Proc Natl Acad Sci USA 71:1056–1060, 1974.

9. Bilodeau-Goeseels S, Schultz GA. Changes in ribosomal ribonucleic acidcontent within in vitro-produced bovine embryos. Biol Reprod 56: 1323–1329, 1997.

10. Birney E, Kumar S, Krainer AR. Analysis of the RNA-recognitionmotif and RS and RGG domains: conservation in metazoan pre-mRNAsplicing factors. Nucleic Acids Res 21: 5803–5816, 1993.

11. Brambrink T, Wabnitz P, Halter R, Klocke R, Carnwath J, Kues W,Wrenzycki C, Paul D, Niemann H. Application of cDNA arrays tomonitor mRNA profiles in single preimplantation mouse embryos. Bio-techniques 33: 376–378, 380, 382–375, 2002.

12. Braude P, Bolton V, Moore S. Human gene expression first occursbetween the four- and eight-cell stages of preimplantation development.Nature 332: 459–461, 1988.

13. Cabot RA, Hannink M, Prather RS. CRM1-mediated nuclear export ispresent during porcine embryogenesis, but is not required for earlycleavage. Biol Reprod 67: 814–819, 2002.

14. Campbell L, Hardwick KG. Analysis of Bub3 spindle checkpoint func-tion in Xenopus egg extracts. J Cell Sci 116: 617–628, 2003.

15. Carayannopoulos MO, Schlein A, Wyman A, Chi M, KeembiyehettyC, Moley KH. GLUT9 is differentially expressed and targeted in thepreimplantation embryo. Endocrinology 145: 1435–1443, 2004.

16. Christians E, Campion E, Thompson EM, Renard JP. Expression ofthe HSP 70.1 gene, a landmark of early zygotic activity in the mouseembryo, is restricted to the first burst of transcription. Development 121:113–122, 1995.

17. Clegg KB, Piko L. Poly(A) length, cytoplasmic adenylation and synthesisof poly(A)� RNA in early mouse embryos. Dev Biol 95: 331–341, 1983.

18. Cleveland WS. Robust locally weighted regression and smoothing scat-terplots. J Am Stat Assoc 368: 829–836, 1979.

19. Conaghan J, Handyside AH, Winston RM, Leese HJ. Effects ofpyruvate and glucose on the development of human preimplantationembryos in vitro. J Reprod Fertil 99: 87–95, 1993.

20. Conover JC, Temeles GL, Zimmermann JW, Burke B, Schultz RM.Stage-specific expression of a family of proteins that are major products ofzygotic gene activation in the mouse embryo. Dev Biol 144: 392–404,1991.

21. Dalbies-Tran R, Mermillod P. Use of heterologous complementary DNAarray screening to analyze bovine oocyte transcriptome and its evolutionduring in vitro maturation. Biol Reprod 68: 252–261, 2003.

22. Davis W Jr, De Sousa PA, Schultz RM. Transient expression oftranslation initiation factor eIF-4C during the 2-cell stage of the preim-plantation mouse embryo: identification by mRNA differential display andthe role of DNA replication in zygotic gene activation. Dev Biol 174:190–201, 1996.

23. Davis W Jr, Schultz RM. Developmental change in TATA-box utiliza-tion during preimplantation mouse development. Dev Biol 218: 275–283,2000.

24. De Rycke M, Liebaers I, Van Steirteghem A. Epigenetic risks related toassisted reproductive technologies: risk analysis and epigenetic inheri-tance. Hum Reprod 17: 2487–2494, 2002.

25. Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC,Lempicki RA. DAVID: Database for Annotation, Visualization, andIntegrated Discovery. Genome Biol 4: P3, 2003.

26. Devreker F, Englert Y. In vitro development and metabolism of thehuman embryo up to the blastocyst stage. Eur J Obstet Gynecol ReprodBiol 92: 51–56, 2000.

27. Donnison M, Pfeffer PL. Isolation of genes associated with developmen-tally competent bovine oocytes and quantitation of their levels duringdevelopment. Biol Reprod 71: 1813–1821, 2004.

28. Dozortsev D, Rybouchkin A, De Sutter P, Qian C, Dhont M. Humanoocyte activation following intracytoplasmic injection: the role of thesperm cell. Hum Reprod 10: 403–407, 1995.

29. Eckert J, Niemann H. In vitro maturation, fertilization and culture toblastocysts of bovine oocytes in protein-free media. Theriogenology 43:1211–1225, 1995.

30. Evsikov AV, de Vries WN, Peaston AE, Radford EE, Fancher KS,Chen FH, Blake JA, Bult CJ, Latham KE, Solter D, Knowles BB.Systems biology of the 2-cell mouse embryo. Cytogenet Genome Res 105:240–250, 2004.

31. Fedoriw AM, Engel NI, Bartolomei MS. Genomic imprinting: antago-nistic mechanisms in the germ line and early embryo. Cold Spring HarbSymp Quant Biol 69: 39–45, 2004.

32. Fornerod M, Ohno M. Exportin-mediated nuclear export of proteins andribonucleoproteins. Results Probl Cell Differ 35: 67–91, 2002.

33. Ge H, Walhout AJ, Vidal M. Integrating “omic” information: a bridgebetween genomics and systems biology. Trends Genet 19: 551–560, 2003.

34. Hamatani T, Carter MG, Sharov AA, Ko MS. Dynamics of global geneexpression changes during mouse preimplantation development. Dev Cell6: 117–131, 2004.

35. Hao Z, Stoler MH, Sen B, Shore A, Westbrook A, Flickinger CJ, HerrJC, Coonrod SA. TACC3 expression and localization in the murine eggand ovary. Mol Reprod Dev 63: 291–299, 2002.

36. Harris MA, Clark J, Ireland A, Lomax J, Ashburner M, Foulger R,Eilbeck K, Lewis S, Marshall B, Mungall C, Richter J, Rubin GM,Blake JA, Bult C, Dolan M, Drabkin H, Eppig JT, Hill DP, Ni L,Ringwald M, Balakrishnan R, Cherry JM, Christie KR, CostanzoMC, Dwight SS, Engel S, Fisk DG, Hirschman JE, Hong EL, Nash RS,Sethuraman A, Theesfeld CL, Botstein D, Dolinski K, Feierbach B,Berardini T, Mundodi S, Rhee SY, Apweiler R, Barrell D, Camon E,Dimmer E, Lee V, Chisholm R, Gaudet P, Kibbe W, Kishore R,Schwarz EM, Sternberg P, Gwinn M, Hannick L, Wortman J, Ber-riman M, Wood V, de la Cruz N, Tonellato P, Jaiswal P, Seigfried T,White R. The Gene Ontology (GO) database and informatics resource.Nucleic Acids Res 32: D258-D261, 2004.

37. Herwig R, Aanstad P, Clark M, Lehrach H. Statistical evaluation ofdifferential expression on cDNA nylon arrays with replicated experiments.Nucleic Acids Res 29: E117, 2001.

38. Hoshino S, Miyazawa H, Enomoto T, Hanaoka F, Kikuchi Y, KikuchiA, Ui M. A human homologue of the yeast GST1 gene codes for aGTP-binding protein and is expressed in a proliferation-dependent mannerin mammalian cells. EMBO J 8: 3807–3814, 1989.

39. Hosoda N, Lejeune F, Maquat LE. Evidence that poly(A) bindingprotein C1 binds nuclear pre-mRNA poly(A) tails. Mol Cell Biol 26:3085–3097, 2006.

40. Kageyama S, Liu H, Nagata M, Aoki F. The role of ETS transcriptionfactors in transcription and development of mouse preimplantation em-bryos. Biochem Biophys Res Commun 344: 675–679, 2006.

41. Kaneko KJ, Cullinan EB, Latham KE, DePamphilis ML. Transcriptionfactor mTEAD-2 is selectively expressed at the beginning of zygotic geneexpression in the mouse. Development 124: 1963–1973, 1997.

42. Kaneko KJ, DePamphilis ML. Regulation of gene expression at thebeginning of mammalian development and the TEAD family of transcrip-tion factors. Dev Genet 22: 43–55, 1998.

43. Ko MS, Kitchen JR, Wang X, Threat TA, Wang X, Hasegawa A, SunT, Grahovac MJ, Kargul GJ, Lim MK, Cui Y, Sano Y, Tanaka T,Liang Y, Mason S, Paonessa PD, Sauls AD, DePalma GE, Sharara R,Rowe LB, Eppig J, Morrell C, Doi H. Large-scale cDNA analysisreveals phased gene expression patterns during preimplantation mousedevelopment. Development 127: 1737–1749, 2000.

44. Kocabas AM, Crosby J, Ross PJ, Otu HH, Beyhan Z, Can H, TamWL, Rosa GJ, Halgren RG, Lim B, Fernandez E, Cibelli JB. Thetranscriptome of human oocytes. Proc Natl Acad Sci USA 103: 14027–14032, 2006.



45. Kues WA, Niemann H. The contribution of farm animals to humanhealth. Trends Biotechnol 22: 286–294, 2004.

46. Latham KE, Rambhatla L, Hayashizaki Y, Chapman VM. Stage-specific induction and regulation by genomic imprinting of the mouseU2afbp-rs gene during preimplantation development. Dev Biol 168: 670–676, 1995.

47. Lazzari G, Wrenzycki C, Herrmann D, Duchi R, Kruip T, NiemannH, Galli C. Cellular and molecular deviations in bovine in vitro-producedembryos are related to the large offspring syndrome. Biol Reprod 67:767–775, 2002.

48. Lin Y, Page DC. Dazl deficiency leads to embryonic arrest of germ celldevelopment in XY C57BL/6 mice. Dev Biol 288: 309–316, 2005.

49. McGrath SA, Esquela AF, Lee SJ. Oocyte-specific expression ofgrowth/differentiation factor-9. Mol Endocrinol 9: 131–136, 1995.

50. Niemann H, Wrenzycki C. Alterations of expression of developmentallyimportant genes in preimplantation bovine embryos by in vitro cultureconditions: implications for subsequent development. Theriogenology 53:21–34, 2000.

51. Parisi T, Beck AR, Rougier N, McNeil T, Lucian L, Werb Z, Amati B.Cyclins E1 and E2 are required for endoreplication in placental trophoblastgiant cells. EMBO J 22: 4794–4803, 2003.

52. Pennetier S, Uzbekova S, Guyader-Joly C, Humblot P, Mermillod P,Dalbies-Tran R. Genes preferentially expressed in bovine oocytes re-vealed by subtractive and suppressive hybridization. Biol Reprod 73:713–720, 2005.

53. Powell K. Fertility treatments: seeds of doubt. Nature 422: 656–658,2003.

54. Riley JK, Moley KH. Glucose utilization and the PI3-K pathway: mech-anisms for cell survival in preimplantation embryos. Reproduction 131:823–835, 2006.

55. Ruddock NT, Wilson KJ, Cooney MA, Korfiatis NA, Tecirlioglu RT,French AJ. Analysis of imprinted messenger RNA expression duringbovine preimplantation development. Biol Reprod 70: 1131–1135, 2004.

56. Said T, Agarwal A, Grunewald S, Rasch M, Baumann T, Kriegel C,Li L, Glander HJ, Thomas AJ Jr, Paasch U. Selection of nonapoptoticspermatozoa as a new tool for enhancing assisted reproduction outcomes:an in vitro model. Biol Reprod 74: 530–537, 2006.

57. Salpekar A, Huntriss J, Bolton V, Monk M. The use of amplified cDNAto investigate the expression of seven imprinted genes in human oocytesand preimplantation embryos. Mol Hum Reprod 7: 839–844, 2001.

58. Sirard MA, Dufort I, Vallee M, Massicotte L, Gravel C, Reghenas H,Watson AJ, King WA, Robert C. Potential and limitations of bovine-specific arrays for the analysis of mRNA levels in early development:preliminary analysis using a bovine embryonic array. Reprod Fertil Dev17: 47–57, 2005.

59. Solter D, Hiiragi T, Evsikov AV, Moyer J, De Vries WN, Peaston AE,Knowles BB. Epigenetic mechanisms in early mammalian development.Cold Spring Harb Symp Quant Biol 69: 11–17, 2004.

60. Stromberg B, Dahlquist G, Ericson A, Finnstrom O, Koster M,Stjernqvist K. Neurological sequelae in children born after in-vitrofertilisation: a population-based study. Lancet 359: 461–465, 2002.

61. Thompson JG, Kind KL, Roberts CT, Robertson SA, Robinson JS.Epigenetic risks related to assisted reproductive technologies: short- andlong-term consequences for the health of children conceived throughassisted reproduction technology: more reason for caution? Hum Reprod17: 2783–2786, 2002.

62. Thompson JG, McNaughton C, Gasparrini B, McGowan LT, TervitHR. Effect of inhibitors and uncouplers of oxidative phosphorylationduring compaction and blastulation of bovine embryos cultured in vitro. JReprod Fertil 118: 47–55, 2000.

63. Tung JY, Rosen MP, Nelson LM, Turek PJ, Witte JS, Cramer DW,Cedars MI, Pera RA. Variants in Deleted in AZoospermia-Like (DAZL)are correlated with reproductive parameters in men and women. HumGenet 118: 730–740, 2006.

64. Vallee M, Gravel C, Palin MF, Reghenas H, Stothard P, Wishart DS,Sirard MA. Identification of novel and known oocyte-specific genes usingcomplementary DNA subtraction and microarray analysis in three differ-ent species. Biol Reprod 73: 63–71, 2005.

65. Van Blerkom J. Mitochondria in human oogenesis and preimplantationembryogenesis: engines of metabolism, ionic regulation and developmen-tal competence. Reproduction 128: 269–280, 2004.

66. Waclaw RR, Chatot CL. Patterns of expression of cyclins A, B1, D, Eand cdk 2 in preimplantation mouse embryos. Zygote 12: 19–30, 2004.

67. Wang J, Mayernik L, Armant DR. Integrin signaling regulates blasto-cyst adhesion to fibronectin at implantation: intracellular calcium tran-sients and vesicle trafficking in primary trophoblast cells. Dev Biol 245:270–279, 2002.

68. Wang Q, Latham KE. Requirement for protein synthesis during embry-onic genome activation in mice. Mol Reprod Dev 47: 265–270, 1997.

69. Wang QT, Piotrowska K, Ciemerych MA, Milenkovic L, Scott MP,Davis RW, Zernicka-Goetz M. A genome-wide study of gene activityreveals developmental signaling pathways in the preimplantation mouseembryo. Dev Cell 6: 133–144, 2004.

70. Whitworth KM, Agca C, Kim JG, Patel RV, Springer GK, Bivens NJ,Forrester LJ, Mathialagan N, Green JA, Prather RS. Transcriptionalprofiling of pig embryogenesis by using a 15-K member unigene setspecific for pig reproductive tissues and embryos. Biol Reprod 72: 1437–1451, 2005.

71. Wrenzycki C, Herrmann D, Lucas-Hahn A, Gebert C, Korsawe K,Lemme E, Carnwath JW, Niemann H. Epigenetic reprogrammingthroughout preimplantation development and consequences for assistedreproductive technologies. Birth Defects Res C Embryo Today 75: 1–9,2005.

72. Wrenzycki C, Herrmann D, Lucas-Hahn A, Korsawe K, Lemme E,Niemann H. Messenger RNA expression patterns in bovine embryosderived from in vitro procedures and their implications for development.Reprod Fertil Dev 17: 23–35, 2005.

73. Wrenzycki C, Herrmann D, Lucas-Hahn A, Lemme E, Korsawe K,Niemann H. Gene expression patterns in in vitro-produced and somaticnuclear transfer-derived preimplantation bovine embryos: relationship tothe large offspring syndrome? Anim Reprod Sci 82–83: 593–603, 2004.

74. Wrenzycki C, Herrmann D, Niemann H. Timing of blastocyst expan-sion affects spatial messenger RNA expression patterns of genes in bovineblastocysts produced in vitro. Biol Reprod 68: 2073–2080, 2003.

75. Wu X, Wang P, Brown CA, Zilinski CA, Matzuk MM. Zygote arrest 1(Zar1) is an evolutionarily conserved gene expressed in vertebrate ovaries.Biol Reprod 69: 861–867, 2003.

76. Yu J, Deng M, Medvedev S, Yang J, Hecht NB, Schultz RM. Trans-genic RNAi-mediated reduction of MSY2 in mouse oocytes results inreduced fertility. Dev Biol 268: 195–206, 2004.

77. Zuccotti M, Boiani M, Ponce R, Guizzardi S, Scandroglio R, GaragnaS, Redi CA. Mouse Xist expression begins at zygotic genome activationand is timed by a zygotic clock. Mol Reprod Dev 61: 14–20, 2002.



Silencing CENPF in bovine preimplantation embryo induces arrest at 8-cell stage

Documents