developmental capacity of porcine nuclear transfer embryos correlate with levels of...

MOLECULAR REPRODUCTION AND DEVELOPMENT 75:766–776 (2008)

Developmental Capacity of Porcine NuclearTransfer Embryos Correlate With Levels ofChromatin-Remodeling Transcripts in Donor CellsLUCA MAGNANI,1 KIHO LEE,1 WILLIAM L. FODOR,2 ZOLTAN MACHATY,1 AND RYAN A. CABOT1*1Department of Animal Sciences, Purdue University, West Lafayette, Indiana2University of Connecticut, Center for Regenerative Medicine, Storrs, Connecticut

ABSTRACT Somatic cell nuclear transfer (SCNT)still retains important limitations. Impaired epigeneticreprogramming is considered responsible for alteredgene expression and developmental failure in SCNT-derived embryos. After nuclear transfer the donor cellnucleus undergoes extensive changes in gene expres-sion that involve epigenetic modifications and chro-matin remodeling. We hypothesized that SNF2-typeATP-dependent chromatin factors contribute to epige-netic reprogramming and the relative amount of thesefactors in the donor cell affects developmental potentialof the reconstructed embryos. In order to test thishypothesis, we assessed the relative amount of SNF2-type ATPases (Brahma, Brg1, SNF2H, SNF2L, CHD3,and CHD5) in three different donor cells as well asin porcine metaphase II oocytes. We performed SCNTwith fetal fibroblast cells, olfactory bulb (OB) progen-itor cells, and porcine skin originating sphere stem cells(PSOS). We found that OB-NT embryos and PSOS-NTembryos resulted in a higher morulae/blastocysts ratioas compared to fibroblast-NT embryos (23.53%,16.98%, and 11.63%, respectively; P<0.05). Fibro-blast cells contained a significantly higher amount ofSNF2L and CHD3 transcripts while Brg1 and SNF2Hwere the most expressed transcripts in all the celllines analyzed. Metaphase II oocyte expression profileappeared to be unique compared to the cell linesanalyzed. This work supports our hypothesis that anarray of chromatin-remodeling proteins on donor cellsmay influence the chromatin structure, effect epige-netic reprogramming, and developmental potential.Mol. Reprod. Dev. 75: 766–776, 2008.� 2008 Wiley-Liss, Inc.

Key Words: embryo; chromatin; nuclear transfer

INTRODUCTION

Somatic cell nuclear transfer (SCNT) proved to be aturning point in the field of biology. After fusion with anenucleated oocyte, a differentiated cell can reconfigureits genetic program and acquire totipotent character-istics. Unfortunately these events happen only with lowfrequency; most of the reconstructed embryos fail toenter a normal developmental program and terminate

development or present gross defects after birth (Younget al., 1998). It is theorized that a major limitation ofSCNT resides in failure in epigenetic reprogrammingin the donor nucleus. Unique genetic programs areestablished in cells as they undergo differentiation (Ganet al., 2007). Epigenetic modifications to chromatin arekey to this process. Changes in DNA methylation,histone methylation, and histone acetylation are tightlylinked to the transcriptional state of a gene (Li, 2002).

During SCNT epigenetic marks need to be remodeledto reprogram the donor cell to behave as a totipotent,single celled embryo. Cloned embryos have been foundto possess epigenetic aberrations such as aberrant DNAmethylation (Wakayama et al., 2000; Dean et al., 2001,2005; Ohgane et al., 2001; Kang et al., 2001a,b,c, 2003;Hiiragi and Solter, 2005; Han et al., 2006). The patternof DNA methylation on satellite DNA resulted wasshown to be altered in bovine embryos produced bySCNT (Kang et al., 2001a,b,c). These SCNT-derivedembryos were found to possess a methylation patternsimilar to that of the donor nucleus, not a stage matchedembryo produced by in vitro fertilization. Studies in thepig showed similar results in which SCNT-derivedembryos possessed a perturbed DNA methylationpattern in the satellite repeats PRE-1 SINE element(Kang et al., 2003). In addition, further aberrationswhere found in the methylation pattern of discretegenomic positions in healthy adult pigs produced bySCNT as compared to control animals (Archer et al.,2003). Another line of studies showed that bothacetylation and methylation status of the lysine 9residue of histone H3 are perturbed in SCNT-derivedbovine embryos (Santos and Reik, 2003).

Besides covalent chemical modifications to DNA andhistone proteins, the chromatin structure can be alteredby repositioning of nucleosomes throughout the genome

� 2008 WILEY-LISS, INC.

William L. Fodor’s present address is ViaCell, Inc., Cambridge, MA02142.

*Correspondence to: Ryan A. Cabot, Department of Animal Sciences,Purdue University, 915 West State Street, West Lafayette, IN 47907.E-mail: [email protected]

Received 23 May 2007; Accepted 3 July 2007Published online 1 February 2008 in Wiley InterScience(www.interscience.wiley.com).DOI 10.1002/mrd.20818

(Kassabov and Vartholomew, 2003; Langst and Becker,2004; Kireeva and Kashlev, 2005; Saha et al., 2006).SNF2-type ATP-dependent chromatin-remodeling com-plexes (from Sucrose NonFermenting 2, SNF2) typicallyexist as multi-subunit protein complexes that utilize theenergy from ATP hydrolysis to reposition nucleosomeson chromatin. The core catalytic subunit of these com-plexes is an ATPase of the SNF2-type. Twenty-nineputative SNF2-type ATPases have been identified in thehuman genome; many of these gene possess orthologsin phylogenetically diverse vertebrate species (Linderet al., 2004). Some of these SNF2-type ATPases havebeen characterized biochemically and are known tofunction as the catalytic subunit of specific chromatin-remodeling complexes. A single SNF2-type ATPase canfunction as the catalytic subunit in multiple chromatin-remodeling complexes; in some cases multiple SNF2-type ATPases can function as the catalytic subunit for asingle complex (Becker and Horz, 2002). It has also beenshown that many SNF2-type ATPases have a tissue-specific and developmentally regulated expressionpattern (Bultman and Magnuson, 2000; Lusser andKadonaga, 2003; Flaus and Owen-Hughes, 2004; Linderet al., 2004; Saha et al., 2006).

Four distinct classes of SNF2-type ATP-dependentchromatin-remodeling complexes have been character-ized biochemically. In the case of mammalian SWI/SNFchromatin-remodeling complexes (from the foundingmembers found in yeast, the switch and sucrose non-fermenting mutants), five distinct chromatin-remodel-ing complexes have been identified (Sudarsanam andWinston, 2000; Ulyanova and Schnitzler, 2005). Eachcomplex has a unique constitution of SWI/SNF subunits.SWI/SNF complexes possess one of the two SNF2-typeATPases as their catalytic subunit, either Brahma(Brm) or Brahma related gene 1 (Brg1). These twoproteins are 75% identical at the amino acid level(Bultman and Magnuson, 2000). In some SWI/SNFcomplexes, for example, the human Brahma associatedfactor (BAF) complex, either Brg1 or Brm is found asthe catalytic subunit. In the human PBAF and hBrmcomplex only Brm is found; in Brg1 complex I and Brg1complex II, only Brg1 has been identified. Some addi-tional subunits, like sucrose nonfermenting mutant 5(SNF5) and Brahma associate factor 53 (BAF53), havebeen found in all five biochemically characterized SWI/SNF complexes (Klochendler-Yeivin et al., 2002). Othersubunits are unique to only a subset of the identifiedcomplexes. Only PBAF complexes contain the BAF180subunit, while Brg1 complex I and complex II, BAF, andhBrm complexes contain the BAF250 subunit.

Studies involving knock-out mice reveal functionaldifferences between specific SNF2-type ATPases. TheBrg1 homozygous knock-out is lethal with embryosfailing to develop beyond pre-implantation stages.Blastocyst outgrowth studies revealed that neither theinner cell mass nor the trophectoderm were able tosurvive in culture, although Brg1 is not necessary forcell viability (Bultman and Magnuson, 2000). Furtherstudies using conditional knock-out strategies revealed

that Brg1 deletion resulted in compromised expressionof 30% of zygotic genes (Bultman et al., 2006). Brmknock-outs showed only a mild overgrowth phenotype(Reyes and Yaniv, 1998). SWI/SNF complexes have beenshown to interact with acetylated residues on histonetails and have been involved in several processesranging from cell cycle control to differentiation (Reyesand Yaniv, 1998; Huang et al., 2002; Gebuhr et al., 2003;Coisy and Dantonel, 2004; Lickert and Bruneau, 2004;Wang et al., 2004) and its distribution in cell types isquite interesting: after a vast screening of humantissues Reisman et al. (2005) came to the conclusionthat Brg1 was expressed preponderantly in undiffer-entiated tissues, where cells were undergoing continu-ous proliferation. Brahma, on the other end, was highlyexpressed in fully differentiated tissues, where cellswere preferably under cell cycle arrest.

Mammalian ISWI complexes, from the foundingmember in yeast imitation switch (ISWI) mutants,contain one of two ATPases, either SNF2H or SNF2L;the former is found in the ACF and CHRAC complexes,while the latter is part of the NURF complex (Ito et al.,1997; Varga-Weisz et al., 1997; Kaji and Kudo, 2004).The ISWI complexes have been implicated in several keyrole such as transcriptional regulation of RNA polymer-ase II (Tsukiyama et al., 1995; Mellor and Morillon,2004), RNA polymerase I (Santoro et al., 2002),chromatin assembly (Deuring et al., 2000), and repli-cation (Collins et al., 2002). Furthermore, a knock-outmouse model showed SNF2H to be necessary for embryoviability (Stopka and Skoultchi, 2003). The ISWIcomplexes have been related to epigenetic changes indifferentiating cells: SNF2L is part of the mammalianNURF complex that is prevalent in the brain where itpromotes the in vitro terminal differentiation of neurons(Barak et al., 2003); it also participates in terminaldifferentiation of granulosa cells in the ovary (Lazzaroet al., 2006). ISWI ATPases have a characteristicexpression profile: SNF2H was found preferentially inproliferating cell types while SNF2L was necessary tomaintain a differentiated phenotype (Lazzaro andPicketts, 2001).

The CHD chromo-domain protein family is involved inthe regulation of gene expression in eukaryotes duringdevelopment (Koonin and Lucchesi, 1995; Jones andSingh, 2000; Eissenberg, 2001). Mammalian Mi-2complexes utilize Mi2a–b, known as CHD3 and CHD4ATPases in the human NuRD complex, as their catalyticsubunit (Bowen et al., 2004; Feng, 2003). In human, sixgenes have been identified as putative CHD ATPases(CHD1, CHD2, CHD3, CHD4, CHD5, and CHD7).CHD5 is closely related to CHD3 and CHD4 and ispreferentially expressed in neuronal tissues (Thompsonet al., 2003). The fourth family is composed by the Ino80/Swr family. This group of chromatin-remodeling com-plexes utilizes a slightly different SNF2-type ATPasewith a longer SNF2 domain. The mechanisms of theremodeling are also quite different; the complexes areable to substitute histone proteins from nuclesomes(reviewed in Morrison, 2006).

Molecular Reproduction and Development

CHROMATIN REMODELING IN PORCINE EMBRYOS 767

Because of the low developmental potential seen inSCNT-derived embryos and the evidence that defectsin epigenetic reprogramming by the oocyte cytoplasm inthe reconstructed embryo contributes to compromisedembryo development, we hypothesized that the chro-matin configuration in the donor cell may contribute tothe efficiency of nuclear reprogramming followingSCNT, and therefore affect the developmental potentialof SCNT-derived embryos. We also hypothesized thatdifferent amounts of SNF2-type ATPases may representa key point in chromatin regulation and this maycorrelate with the efficiency in reprogramming follow-ing SCNT. The objectives of this study were to determinehow the relative abundance of six SNF2-type ATPasescompared between metaphase II-arrested oocytes andthree donor cell types and to determine if the differencein SNF2-type ATPases correlated with the in vitrodevelopmental potential of SCNT-derived porcineembryos.

MATERIALS AND METHODS

Oocyte and Embryo Production

All chemicals were obtained from Sigma ChemicalCompany (St. Louis, MO) unless stated otherwise.Porcine (Sus scrofa) ovaries from pre-pubertal giltswere collected at a local slaughterhouse and transportedto the laboratory in an insulated container at 378C.Antral follicles between 3 and 6 mm in diameter wereaspirated manually with a disposable 10-cc syringeand an 18-gauge needle. Follicular fluid was pooled andallowed to settle by gravity. Cumulus-oocyte complexes(COCs) were resuspended in Hepes-buffered Tyrode’smedium containing 0.01% polyvinyl alcohol (PVA)(Abeydeera et al., 1998). Under a dissecting microscope,COCs with multiple layers of intact cumulus cells wereselected for the experiments.

In Vitro Maturation

Fifty to 75 COCs were placed in 500ml of tissue culturemedium 199 (TCM-199; Gibco BRL, Grand Island, NY)containing 0.14% PVA, 10 ng/ml epidermal growthfactor (EGF), 0.57 mM cysteine, 0.5 IU/ml porcineFSH, and 0.5 IU/ml ovine LH. COCs were matured for44 hr at 398C and 5% CO2 in air, 100% humidity(Abeydeera et al., 1998). Matured COCs were thenvortexed in 0.1% hyaluronidase in Hepes-bufferedTyrode’s medium containing 0.01% PVA for 4 min toremove the cumulus cells.

Porcine Fibroblast, PSOS Stem Cell, andOlfactory Bulb Progenitor Cell

Isolation and Culture

All experiments involving animals were conductedaccording to protocols approved by Purdue UniversityAnimal Care and Use Committee. Porcine fetal fibro-blast cells were isolated from porcine fetuses at Day 35 ofgestation. Briefly, fetuses were removed from the uterusof an artificially inseminated sow and washed inDulbecco’s Phosphate Buffered Saline (DPBS). After

removing the head and the internal organs the fetuseswere cut into approximately 1 mm3 pieces in DPBS.The tissue pieces were then seeded in a 25 cm2 tissueculture flask containing 2 ml of Dulbecco’s ModifiedEagle Medium (DMEM) supplemented with 15% fetalbovine serum and 1% penicillin/streptomycin. After3 days the tissue explants were removed by rinsing theflask with DPBS. The attached fibroblast cells werecultured until they reached confluency. To passage cellstrypsinization using 0.1% trypsin in 0.02% EDTA wasused. Porcine skin originating sphere stem cells (PSOS)stem cells were isolated as described previously, withminor modifications (Dyce et al., 2004). Briefly, skin wascollected from porcine fetuses at Day 70 of gestation. Theskin tissue was cut into small (�1 mm3) pieces andwashed three times in Hanks Balanced Salt Solution(HBSS), then digested with 0.1% trypsin for 40 min at378C. After the digestion, the sample was treated with0.1% DNase for 1 min at room temperature and washedin HBSS. The sample was washed in culture medium(DMEM-F12 (3:1), 1 mg/ml fungizone, 1% penicillin/streptomycin) with 10% fetal bovine serum and thenwashed twice with serum-free culture medium. Thetissue pieces were dissociated mechanically in themedium, poured through a 40 mm cell strainer (Falcon)and spun at 168g for 10 min. The collected cell sus-pension was resuspended in 10 ml culture mediumcontaining B27, 20 ng/ml EGF and 40 ng/ml basicfibroblast growth factor (bFGF). The cells were culturedin 25 cm2 tissue culture flasks. Small floating cellspheres started to form after 3–7 days. The floatingclusters of cells were passaged after every 7–10 days.OB progenitor cells were isolated from porcine fetusesat Day 45 of gestation (Pagano et al., 2000). The OBswere removed from the fetuses and the ensheathing gliaouter layer was dissected away from the rest of the bulb.Cells were dissociated with liberase blendzyme 3 (RocheApplied Science, Indianapolis, IN) and cultured inserum-free neural stem cell (NSC) medium (CellGroComplete medium (Mediatech, Inc.; Herndon, VA))supplemented with 4 ng/ml bFGF, 10 ng/ml EGF, 1XN2 supplement (Invitrogen Corporation; Carlsbad, CA)and 1% penicillin/streptomycin. The adherent stem cellpopulation was maintained in NSC medium indefinitely(>1 year). Cells at passage 2 until passage 7 were usedfor the experiments.

Oct-4 Immunocytochemistry

Cells were washed in PBS and fixed in 3.7%paraformaldehyde in PBS for 15 min at room temper-ature. After fixation the cells were washed and perme-ablized in PBS containing 0.1% Triton X-100 for 15 min.Following permeabilization, the cells were washed inPBS containing 0.01% Tween-20 with 2% BSA. Non-specific binding sites were blocked by incubation in thepresence of 0.01% Tween-20 with 2% BSA for 1 hr atroom temperature or overnight at 48C. The cells werethen incubated with mouse monoclonal IgG primaryantibody raised against the Oct-4 protein (Chemicon,dilution 1:100) for 1 hr at room temperature. This was


768 L. MAGNANI ET AL.

followed by incubation with rhodamine-conjugatedrabbit polyclonal anti-mouse secondary antibody(Abcam, 1:200) for 30 min at room temperature. Hoechst33342 (1.2 mg/ml) was used to stain DNA. After washingwith PBS the cells were examined by fluorescencemicroscopy (Nikon ECLIPSE TE2000-U) using theappropriate filters; pictures were taken using a Nikondigital camera (Nikon DS-5M) at 20� magnification.

Nuclear Transfer

Oocytes were enucleated in Hepes-buffered Tyrode’smedium containing 7.5 mg/ml cytochalasin B and 5% ofsucrose. Prior to enucleation, the oocytes were incubatedin the presence of 1.2 mg/ml Hoechst 33342 for 10 min tovisualize the nuclear material. The metaphase platetogether with the adjacent cytoplasm was removedusing a beveled micropipette; a nuclear donor cell wasthen placed under the zona pellucida. To minimize thevariation in oocyte quality on a certain day, all three celltypes were used for the reconstruction of embryoseach day. For fusion, reconstructed oocytes were washedin a calcium-free fusion medium consisting of 300 mMmannitol, 0.2 mM MgSO4, 0.5 mM Hepes, and 0.1%PVA. They were then placed between the electrodes ofan electroporation chamber filled with the fusionmedium. Fusion of the plasma membranes was inducedwith two DC pulses of 1.2 kV/cm for 60 msec each (using aBTX Electro Cell Manipulator, Harvard Apparatus,Holliston, MA); 2 min later the reconstructed oocyteswere transferred into Hepes-buffered Tyrode’s mediumand cultured for 1 hr. The reconstructed oocytes werethen activated by two DC pulses of 1.2 kV/cm for 60 mseceach; the medium for activation consisted of 300 mMmannitol, 0.1 mM CaCl2, 0.1 mM MgSO4, 0.5 mM Hepes,and 0.1 mg/ml BSA. Activated embryos were incubatedin NCSU-23 medium containing 10 mg/ml cycloheximideand 5 mg/ml cytochalasin B for 5 hr. The embryos weresubsequently cultured in NCSU-23 medium containing4 mg/ml BSA for 6 days. At the end of the 6-day cultureperiod, embryos were assessed morphologically and the

number of nuclei per embryo was determined byHoechst staining.

Cloning of Porcine SNF2-Type ATPases

Several porcine ESTs containing sequence datahighly similar to the respective human SNF2 orthologs(>90% identity at the nucleotide level) were employed toclone 400–700 bp sequences for each gene into thePCRII-TOPO vector (Invitrogen). Primers specific forporcine sequences were derived from the followingESTs: Brg1 (BP162626), Brahma (BP158822), SNF2L(BI184017), SNF2H (CK458110), CHD3 (AU296516),and CHD5 (CK462412). The resulting amplicons weresequenced and used to design real time oligonucleotideprimers. Standard curves for each set of oligonucleotideprimers were performed to compare amplificationefficiency, only oligonucleotide primers with similarefficiency were used in real-time PCR experiments.Primers are shown in Table I.

RNA Isolation, Reverse Transcription

Cell lines. RNA was extracted from the three celllines cultured as described for the SCNT procedure.RNA was obtained using RNeasy (Qiagen, Valencia, CA)according to the manufacturer’s instructions. RNA wasanalyzed on agarose gel to assess quality and quantifiedusing ND1000 (Nanodrop, Wilmington, DE). RNAextracted from three pools of cells cultured from threedifferent animals was used in this experiment. Reversetranscription was performed using the Iscript kit (Bio-Rad, Hercules, CA) using 2.4 mg of RNA in each 20 mlreaction. For real-time PCR a master-mix composed of12.5 ml of SybrGreen Master mix (Bio-Rad), 10 ml of 1 mMmixed forward and reverse primer and 2.5 ml of each RTreaction was prepared. Reactions were carried out intriplicates for each gene. For each replicate, cDNA fromone pool of cells was divided to amplify all genes duringthe same PCR run. As a negative control, reversetranscription reactions lacking reverse transcriptasewas used.


TABLE I. Primers for Real-Time PCR

Gene Melting temperature

YWHAG-F 50-TCCATCACTGAGGAAAACTGCTAA-30 78.0YWHAG-R 50-TTTTTCCAACTCCGTGTTTCTCTA-30

GAPDH-F 50-TGCCAGTGAGCTTCCCGTT 82.0GAPDH-R 50-CCCAGAACATCATCCCTGCTTCTABRAHMA-F 50-GGGACCCCACTGCAGAATAAGC-30 84.5BRAHMA-R 50-AATCTGGGGGAGGAGGAAGTTGAG-30

BRG1-F 50-AGACGAAGTGCCTGACGATGAGAC-30 89.0BRG1-R 50-GCTTCCGCTTGGGGTTGC-30

SNF2H-F 50-ATTGTAGGTTGGATGGGCAGACACC-30 82.5SNF2H-R 50-CACCAGCACGTGTGCTTAACATGAA-30

SNF2L-F 50-TGCAGGAGCTCGAATAAGCGT-30 85.5SNF2L-R 50-CGGTGGATGCCTACTTCAGAGA-30

CHD3-F 50-GCCGGCTCCTGTGATAGGTT-30 89.0CHD3-R 50-TTCAAGCTCCTGGAGCAGGC-30

CHD5-F 50-CTCCCCGTGTGACCACAGCACTGAT-30 85.5CHD5-R 50-TGCAGCAGCGTACAGGTCCTTCTG-30


Metaphase II oocytes. For mRNA isolation frommetaphase II oocytes, pools of 100–150 oocytes werecultured for 42–44 hr as described above. RNA extractedfrom three pools of oocytes collected on separate dayswas used in this experiment as experimental replicates.After maturation, oocytes were denuded and washedthree times in PVA-PBS (nuclease free) then lysed in100 ml of lysis buffer (Invitrogen Corporation, Carlsbad,CA). After lysis, mRNA was stored at �808C untilreverse transcription. Reverse transcription was per-formed using the Iscript kit (Bio-Rad). For eachexperimental replicate, cDNA from one pool of oocyteswas divided to amplify all genes during the same PCRrun. As a negative control, reverse transcriptionreactions lacking reverse transcriptase was used.

Quantitative PCR

PCR was performed on a MyiQ single color real-timethermal cycler (Bio-Rad). The following program wasused to perform PCR: 50 initial denaturation at 948Cfollowed by 40 cycles of 5 sec at 948C, 30 sec at 608C, and30 sec at 728C. A melting curve was produced to verifyindividual transcripts. Products were confirmed on anagarose gel; in the initial replicate, transcripts werecloned to verify these exact sequences.

Quantification of Transcript Levels

Cell lines. Transcript levels of all genes werequantified using the comparative threshold cycle (CT)method. For quantifying transcripts the relativeamount of each target gene was determined relative tothe housekeeping gene GAPDH. GAPDH was chosen touse in the quantification method since it had previouslybeen shown to maintain a constant level of expression inthese cell types (Duvigneau et al., 2005).

Metaphase II oocytes. Transcript levels of allgenes were quantified using the comparative thresholdcycle method. For quantifying transcripts the relativeamount of each target gene was determined relative tothe housekeeping gene YWHAG. YWHAG was chosen touse in the quantification method since it had previouslybeen shown to express at a constant level throughoutpre-implantation development (Whitworth et al., 2005);in addition, YWHAG has been used to quantify tran-scripts in porcine oocytes (Magnani and Cabot, 2007).The CT value, or the cycle number during the log-linearphase of the PCR at which time the amount of detectedPCR product rises above background levels was deter-mined for each reaction. Briefly, the CT value for eachtarget gene was subtracted from the CT value for thehousekeeping gene to obtain the change in CT (theDCT).

Statistical Analysis

Developmental data pertaining to the nuclear trans-fer experiment were analyzed by ANOVA and Tukey’spost-test for multiple comparisons; a P< 0.05 was con-sidered significant. For quantitative PCR, 2jDCTj valueswere obtained, imported into SAS and analyzed in GLMone way ANOVA with LSD test for multiple compar-isons; a P-value of <0.05 was considered significant.

RESULTS

OB Progenitor Cells and PSOS Stem CellsContain the Transcription Factor Oct-4

Immunocytochemical analysis of porcine fibroblastcells, PSOS stem cells and OB progenitor cells revealedthat only the PSOS and OB cells expressed Oct-4, amarker of pluripotency (Fig. 1). This result confirms thefinding reported for the OB and PSOS stem cells(Pagano et al., 2000; Dyce et al., 2004).


Fig. 1. PSOS stem cells and OB progenitor cells possess Oct-4protein. PSOS stem cells are positive for pluripotency marker Oct-4when in sphere (Panel A). Oct-4 signal is not lost upon spheredisruption (Panel C). OB progenitor cells are positive for pluripotentmarker Oct-4 (Panel E). Panels B,D,F,H represent nuclear DNAstaining. Fibroblast cells do not possess the pluripotent marker Oct-4(Panel G).


PSOS Stem Cells and OB ProgenitorCells Support Better Development of

SCNT-Derived Embryos In Vitro

Three cell types were used as nuclear donors togenerate porcine embryos using nuclear transfer. Atotal of 161 embryos were reconstructed using PSOSstem cells, 171 embryos from OB cells, and 189 embryosfrom fibroblast cells. Results are summarized inFigure 2. Of the PSOS stem cell-derived embryos,32.9% cleaved and during subsequent culture 5.6%developed to the morula/blastocyst stage. In the OBprogenitor cell group, 19.8% cleaved and 4.7% developedto the morula/blastocysts stage. In the fibroblast group,22.7% cleaved and 2.6% developed to the morula/blastocysts stage. A significantly higher proportion ofembryos reconstructed from PSOS stem cells cleavedcompared to embryos reconstructed with the other celltypes (P< 0.05). In addition, morula/blastocyst forma-tion from PSOS stem cell-derived embryos was signifi-cantly higher than that from fetal fibroblast-derivedembryos (P<0.05); morula/blastocyst formation fromOB cell-derived embryos also tended to be highercompared to embryos produced using porcine fibroblastcells as nuclear donors (P¼ 0.08). The formation ofmorulae/blastocysts per cleaved embryos was the high-est in embryos reconstructed with OB cells (23.5% vs.17.0% using PSOS stem cells and 11.6% using fibro-blasts, P<0.05).

Different Cell Lines Contain Differing Levels ofChromatin-Remodeling ATPase Transcripts

We next investigated the composition of SNF2-typeATPases in each of these donor cell types. Of the sixSNF2-type ATPases analyzed, we found SNF2H andBrg1 transcripts to be the most abundant in PSOS, OB,

and fibroblast cells. SNF2L and CHD3 presentedinteresting differences (Fig. 3). SNF2L transcripts werefound to be fairly abundant in fibroblast cells while OBand PSOS cells had a significantly lower amount(P<0.05). Transcripts for CHD3 were significantlyhigher in PSOS stem cells and fibroblast cells whencompared with OB cells (P<0.05). Brahma and CHD5did not present any significant difference althoughBrahma showed a tendency to be expressed at higherlevel in fibroblast cells compared to OB and PSOS stemcells.

SNF2-Type ATPases Expressionin Metaphase II Oocytes

We next investigated the expression profile of SNF2-type ATPases present in mature porcine oocytes (Fig. 4).We found that metaphase II oocytes are particularly richin SNF2H and CHD5. Oocytes possessed a relativelyhigh amount of Brahma transcripts while SNF2L andCHD3 were in lower abundance when compared to theother SNF2-type ATPases (P<0.05). Brg1 was express-ed at an intermediate level which was not statisticallydifferent from Brahma or SNF2L.

DISCUSSION

The idea that undifferentiated cells support superiorembryonic development when employed as nucleardonors during nuclear transfer is based on the idea thatstem cell-type cells require less nuclear reprogrammingcompared to terminally differentiated cells. This may bedue to a more plastic chromatin status granted by a morepermissive layout of covalent chromatin modifications.The idea is supported by reports in mice, where ES stemcells support term development with an efficiency 3–10 times higher then differentiated somatic cells (Saito


Fig. 2. Specific donor cell types have different developmental potential in vitro. Y-axis represents theactual percentage that developed at each stage. Different superscripts represent statistical difference afterTukey’s multiple comparison post-test (P< 0.05).


et al., 2004; Meissner and Jaenisch, 2006). In domesticanimals, similar conclusions have been drawn usingpluripotent cell types like mesenchymal stem cells (Jinet al., 2007). A recent work, however, reports that amongthe mouse hematopoietic cell lineage, differentiatedcells support higher developmental potential (Sunget al., 2006). This suggests that the link between dif-ferentiation status and developmental potential afternuclear transfer may be more complex.

We hypothesized that the chromatin status in thedonor cell may be of fundamental importance forreprogramming events following nuclear transfer.Chromatin conformation may be dependant not only

on covalent modifications, but also on the set ofchromatin-remodeling complexes present in the cell.These complexes are important cofactors in severalprocesses and work by displacing nucleosomes andallowing docking sites for transcription factors (Beckerand Horz, 2002). In this study, we performed SCNT withthree different cell types and analyzed the transcriptabundance of 6 SNF2-type ATPases present in each ofthese cell types. Higher developmental potential in vitrowas achieved using PSOS and OB cells, two undiffer-entiated cell types. Intriguingly these two cell typesshared some similarities in chromatin-remodeling factorscomposition when compared to metaphase II oocytes.


Fig. 3. Individual cell types used as donor cells in SCNT possess a unique SNF2-type ATPase expressionpattern. Fold differences were calculated using the 2DCT formula relative to the housekeeping gene GAPDHand are reported in the grid. Different subscripts represent significant difference after LSD multiplecomparison post-test (P<0.05). Please note units on Y-axes are expressed in different increments fordifferent genes.


While SNF2H was expressed at high levels in all celltypes analyzed in this work, SNF2L transcripts werepresent in lower abundance in OB progenitor cells andPSOS stem cells as compared to fibroblast cells(P< 0.05). In metaphase II oocytes, SNF2H was themost abundant transcript while SNF2L was present insignificantly lower amounts (P< 0.05). ISWI proteinsSNF2H and SNF2L are the ATPase subunits in threebiochemically characterized complexes: ATP-utilizingchromatin assembly and remodeling complex (ACF),chromatin-accessibility complex (CHRAC), and nucleo-some remodeling factor (NURF) (Tsukiyama and Wu,1995; Tsukiyama et al., 1995; Ito et al., 1997; Varga-Weisz et al., 1997). SNF2H was found prominently in themammalian equivalent of ACF and CHRAC and is foundpredominantly in cell types undergoing proliferation(Lazzaro and Picketts, 2001). SNF2L, on the other hand,was isolated as part of the mammalian NURF complex,where it supports terminal differentiation in neuronalcells and granulosa cells in vitro (Lazzaro and Picketts,2001; Lazzaro et al., 2006). Metaphase II oocytes need tosustain a fast round of cell division after fertilizationwhile preserving pluripotency, this may explain theparticular distribution of SNF2H and SNF2L observedin our experiment.

Fibroblast cells tended to have a higher amount ofBrahma transcript than the other cell types, althoughnot significantly. Brahma protein was shown to beaccumulated in G0, while it was found to be almostundetectable after serum addition (Muchardt andYaniv, 2001). This may have increased our experimentalvariance since our cells were not synchronized. Brahmaand Brg1 are 75% identical at the amino acid level buttheir developmental requirements are different. WhileBrahma null mice are viable with only minor pheno-types, Brg1 knock-out mice are lethal at pre-implanta-tion stages (Bultman and Magnuson, 2000).

Of the five SWI/SNF complexes isolated in human,two of them can use alternatively Brg1 or Brahma as

ATPase subunit, specifically PBAF and Brg1/Brahma(Eberharter and Becker, 2004). The presence of alter-nate subunits may direct the complex to different targetgenes depending on the identity of the ATPase.Pluripotent cells tended to possess a lower amount ofBrahma (hence a lower amount of Brahma PBAF andBrahma Brg1/Brahma complex), this in turn may resultin a different set of genes controlled by BAF and Brg1/Brahma as opposed to differentiated cells with higherBrahma expression. We previously showed that over-expression of Brahma RNA but not a dominant negativeversion dramatically reduced embryo development,suggesting that an increased amount of Brahma-basedcomplexes may be detrimental to embryo development(Magnani and Cabot, 2007).

Another similarity between oocytes and cells withhigher developmental potential was found while ana-lyzing CHD3. OB progenitor cells supported higherdevelopment in vitro (P<0.05). Interestingly, CHD3was found to be in low abundance in both oocytes and OBprogenitor cells whereas PSOS possessed a higheramount of CHD3 (P< 0.05). More studies are neededto define the exact composition of the CHD complexin mammals. Future studies will analyze the relativeamount of the remaining CHD ATPases in porcineembryos and donor cells.

Based on our preliminary findings, we propose amodel to explain a possible role of chromatin-remodelingcomplexes during nuclear reprogramming events afternuclear transfer (Fig. 5). Upon differentiation severalepigenetic marks are positioned to modulate geneexpression, to restrict plasticity and commit the cell toa specific genetic program. This may involve chromatin-remodeling factors. Cells may turn on the expression ofSNF2L and Brahma upon differentiation and thischange in the balance of SNF2-type ATPases may leadto a change in the balance of specific chromatin-remodeling complexes with specificity for uniquegenomic targets. Differentiated porcine fetal fibroblasts


Fig. 4. Metaphase II oocytes possess a defined SNF2-type ATPases expression pattern. Fold differenceswere calculated using the 2DCT formula relative to the housekeeping gene YWHAG and are reported in thegrid. Different subscripts represent significant difference after LSD multiple comparison post-test(P< 0.05).


displayed an increase in SNF2L and Brahma. This mayincrease the amount of BAF and NURF complexescontaining Brahma and SNF2L.

We speculate that a particular array of SNF2-typeremodeling factors serves a key role in positioningdonor cell chromatin for efficient reprogramming by theoocytes. Following nuclear transfer the oocyte ispresented with a chromatin structure that is differentfrom that of a sperm cell. The oocyte’s reprogrammingmachinery acts on the donor nuclear material and canremodel the epigenetic state and initiate the correctexpression program. Intriguingly, the pluripotent cellsused as donor cells in our work possess a SNF2-typeATPase expression profile similar to the oocyte. Thismay prompt the chromatin in a more accessible con-formation to the reprogramming factors present in theoocyte. This in consequence may account for the improv-ed reprogramming efficiency.

Our data showed that pluripotent cells possess acharacteristic set of SNF2-type ATPases: both PSOSstem cells and OB cells had similar low expression ofBrahma and SNF2L and both cell types supported asuperior development in vitro once used as donor cell.While the in vitro development of the SCNT-derivedembryos used in our study was low and we have no datademonstrating term development with these donor cellsin our lab, our data demonstrate a link between theexpression pattern of specific chromatin-remodelingfactors and the developmental potential of SCNT-derived embryos. Further studies will characterize therelative abundance of SNF2-type ATPases in in vivo-derived and SCNT-derived pre-implantation embryos toidentify important key players in nuclear reprogram-ming. Moreover, manipulation of the SNF2-type reper-toire in donor cells may be beneficial for nuclear

reprogramming and increase developmental potentialof SCNT-derived embryos.

ACKNOWLEDGMENTS

We thank Dr. Randall S. Prather for providing us withthe YWHAG construct for use in our PCR experiments.Our appreciation to the Indiana Packers Corporation forthe gift of the porcine ovaries used in this study.

REFERENCES

Abeydeera LR, Wang WH, Prather RS, Day BN. 1998. Maturationin vitro of pig oocytes in protein-free culture media: Fertilization andsubsequent embryo development in vitro. Biol Reprod 58:1316–1320.

Archer GS, Dindot S, Friend TH, Walker S, Zaunbrecher G, Lawhorn B,Piedrahita JA. 2003. Hierarchical phenotypic and epigeneticvariation in cloned swine. Biol Reprod 69:430–436.

Barak O, Lazzaro MA, Lane WS, Speicher DW, Picketts DJ,Shiekhattar R. 2003. Isolation of human NURF: A regulator ofEngrailed gene expression. Embo J 22:6089–6100.

Becker PB, Horz W. 2002. ATP-dependent nucleosome remodeling.Annu Rev Biochem 71:247–273.

Bowen NJFN, Kajita M, Wade PA. 2004. Mi-2/NuRD: Multiplecomplexes for many purposes. Biochim Biophys Acta 1677:52–57.

Bultman S, Magnuson T. 2000. A Brg1 Null mutation in the mousereveals functional differences among mammalian SWI/SNF com-plexes. Molecular Cell 6:1287–1295.

Bultman SJ, Gebuhr TC, Pan H, Svoboda P, Schultz RM, Magnuson T.2006. Maternal BRG1 regulates zygotic genome activation in themouse. Genes Dev 20:1744–1754.

Coisy M, Dantonel J-C. 2004. Cyclin-A repression in quiescent cells isassociated with chromatin remodeling of its promoter and requiresBrahma/SNF2(alpha). Molecular Cell 15:43–56.

Collins N, Poot RA, Kukimoto I, Garcia-Jimenez C, Dellaire G, Varga-Weisz PD. 2002. An ACF1-ISWI chromatin-remodeling complex isrequired for DNA replication through heterochromatin. Nat Genet32:627–632.

Dean W, Santos F, Stojkovic M, Zakhartchenko V, Walter J, Wolf E,Reik W. 2001. Conservation of methylation reprogramming inmammalian development: Aberrant reprogramming in clonedembryos. Proc Natl Acad Sci U S A 98:13734–13738.


Fig. 5. Proposed model. Specific SNF-2 type ATPases present in the donor cell affect epigeneticreprogramming and the establishment of totipotency in the reconstructed embryo.


Dean W, Lucifero D, Santos F. 2005. DNA methylation in mammaliandevelopment and disease. Birth Defects Res C Embryo Today 75:98–111.

Deuring R, Fanti L, Armstrong JA, Sarte M, Papoulas O, Prestel M,Daubresse G, Verardo M, Moseley SL, Berloco M, Tsukiyama T, WuC, Pimpinelli S, Tamkun JW. 2000. The ISWI chromatin-remodelingprotein is required for gene expression and the maintenance ofhigher order chromatin structure in vivo. Mol Cell 5:355–365.

Duvigneau JCHR, Groiss S, Gemeiner M. 2005. Quantitative simulta-neous multiplex real-time PCR for the detection of porcine cytokines.J Immunol Methods 306:1–2.

Dyce PW, Zhu H, Craig J, Li J. 2004. Stem cells with multilineagepotential derived from porcine skin. Biochem Biophys Res Commun316:651–658.

Eberharter A, Becker PB. 2004. ATP-dependent nucleosome remodel-ling: Factors and functions. J Cell Sci 117:3707–3711.

Eissenberg J. 2001. Molecular biology of the chromo domain: Anancient chromatin module comes of age. Gene 275:19–29.

Feng QZY. 2003. The NuRD complex: Linking histone modification tonucleosome remodeling. Curr Top Microbiol Immunol 274:269–290.

Flaus A, Owen-Hughes T. 2004. Mechanisms for ATP-Dependentchromatin remodeling: Farewelllto the tuna-can octamer? Curr OpinGenet Dev 14:165–173.

Gan QYT, McDonald OG, Owens GK. 2007. Concise review: Epigeneticmechanisms contribute to pluripotency and cell lineage determina-tion of embryonic stem cells. Stem Cells 25:2–9.

Gebuhr TC, Kovalev GI, Bultman S, Godfrey V, Su L, Magnuson T.2003. The role of Brg1, a catalytic subunit of mammalian chromatin-remodeling complexes, in T cell development. J Exp Med 198:1937–1949.

Han YM, Kim SH, Kang YK. 2006. Analysis of DNA methylationprofiles in preimplantation embryos using bisulfite mutagenesis.Methods Mol Biol 325:251–260.

Hiiragi T, Solter D. 2005. Reprogramming is essential in nucleartransfer. Mol Reprod Dev 70:417–421.

Huang M, Qian F, Hu Y, Ang C, Li Z, Wen Z. 2002. Chromatin-remodelling factor BRG1 selectively activates a subset of interferon-alpha-inducible genes. Nat Cell Biol 4:774–781.

Ito T, Bulger M, Pazin MJ, Kobayashi R, Kadonaga JT. 1997. ACF, anISWI-containing and ATP-utilizing chromatin assembly and remod-eling factor. Cell 90:145–155.

Jin HF, Kumar BM, Kim JG, Song HJ, Jeong YJ, Cho SK,Balasubramanian S, Choe SY, Rho GJ. 2007. Enhanced developmentof porcine embryos cloned from bone marrow mesenchymal stemcells. Int J Dev Biol 51:85–90.

Jones DOCI, Singh PB. 2000. Mammalian chromodomain proteins:Their role in genome organisation and expression. Bioessays 2:124–137.

Kaji K, Kudo K. 2004. the mechanism of sperm-oocyte fusion inmammals. Reproduction 127:424–429.

Kang YK, Koo DB, Park JS, Choi YH, Chung AS, Lee KK, Han YM.2001a. Aberrant methylation of donor genome in cloned bovineembryos. Nat Genet 28:173–177.

Kang YK, Koo DB, Park JS, Choi YH, Kim HN, Chang WK, Lee KK,Han YM. 2001b. Typical demethylation events in cloned pig embryos.Clues on species-specific differences in epigenetic reprogramming ofa cloned donor genome. J Biol Chem 276:39980–39984.

Kang YK, Koo DB, Park JS, Choi YH, Lee KK, Han YM. 2001c.Differential inheritance modes of DNA methylation between euchro-matic and heterochromatic DNA sequences in ageing fetal bovinefibroblasts. FEBS Lett 498:1–5.

Kang YK, Lee KK, Han YM. 2003. Reprogramming DNA methylationin the preimplantation stage: Peeping with Dolly’s eyes. Curr OpinCell Biol 15:290–295.

Kassabov SR, Vartholomew B. 2003. SWI/SNF unwraps, slides, andrewraps the nucleosome. Molecular Cell 11:391–403.

Kireeva M, Kashlev M. 2005. Nature of the nucleosomal barrier to RNApolimerase II. Molecular Cell 18:97–108.

Klochendler-Yeivin A, Muchardt C, Yaniv M. 2002. SWI/SNF chroma-tin remodeling and cancer. Curr Opin Genet Dev 12:73–79.

Koonin EVZS, Lucchesi JC. 1995. The chromo superfamily: Newmembers, duplication of the chromo domain and possible role in

delivering transcription regulators to chromatin. Nucleic Acids Res23:4229–4233.

Langst G, Becker PB. 2004. Nucleosome remodeling: One mechanism,many phenomena? Biochim Biophys Acta 1677:58–63.

Lazzaro MA, Picketts DJ. 2001. Cloning and characterization of themurine Imitation Switch (ISWI) genes: Differential expressionpatterns suggest distinct developmental roles for Snf2h and Snf2l.J Neurochem 77:1145–1156.

Lazzaro MA, Pepin D, Pescador N, Murphy BD, Vanderhyden BC,Picketts DJ. 2006. The imitation switch protein SNF2L regulatessteroidogenic acute regulatory protein expression during terminaldifferentiation of ovarian granulosa cells. Mol Endocrinol 20:2406–2417.

Li E. 2002. Chromatin modification and epigenetic reprogramming inmammalian development. Nat Rev Genet 3:662–673.

Lickert H, Bruneau BG. 2004. Baf60c is esential for function of BAFchromatin remodelling complexes in heart development. Nature432:107–112.

Linder B, Cabot R, Rupp RAW. 2004. The SNF2 domain protein familyin higher vertebrates displays dynamic expression pattern inxenopus laevis embryos. Gene 326:59–66.

Lusser A, Kadonaga JT. 2003. Chromatin remodeling by ATP-depend-ent molecular machines. BioEssays 25:1192–1200.

Magnani L, Cabot RA. 2007. Developmental arrest induced in cleavagestage porcine embryos following microinjection of mRNA encodingBrahma (Smarca 2), a chromatin remodeling protein. Mol ReprodDev 74:1262–1267.

Meissner A, Jaenisch R. 2006. Mammalian nuclear transfer. Dev Dyn235:2460–2469.

Mellor J, Morillon A. 2004. ISWI complexes in Saccharomycescerevisiae. Biochim Biophys Acta 1677:100–112.

Morrison AJSX. 2006. Chromatin modifications in DNA repair. ResultsProbl Cell Differ 41:109–125.

Muchardt C, Yaniv M. 2001. When the SWI/SNF complex remodels thecell cycle. Oncogene 20:3067–3075.

Ohgane J, Wakayama T, Kogo Y, Senda S, Hattori N, Tanaka S,Yanagimachi R, Shiota K. 2001. DNA methylation variation incloned mice. Genesis 30:45–50.

Pagano SF, Impagnatiello F, Girelli M, Cova L, Grioni E, Onofri M,Cavallaro M, Etteri S, Vitello F, Giombini S, Solero CL, Parati EA.2000. Isolation and characterization of neural stem cells from theadult human olfactory bulb. Stem Cells 18:295–300.

Reisman DN, Sciarrotta J, Bouldin TW, Weissman BE, FunkhouserWK. 2005. The expression of the SWI/SNF ATPase subunits BRG1and BRM in normal human tissues. Appl Immunohistochem MolMorphol 13:66–74.

Reyes JC, Yaniv M. 1998. Altered control of cellular proliferation in theabsence of mammalian brahma. EMBO J 17:6979–6991.

Saha A, Wittmeyer J, Cairns BR. 2006. Chromatin remodelling: Theindustrial revolution of DNA around histones. Nat Rev Mol Cell Biol7:437–447.

Saito S, Liu B, Yokoyama K. 2004. Animal embryonic stem (ES) cells:Self-renewal, pluripotency, transgenesis and nuclear transfer. HumCell 17:107–115.

Santoro R, Li J, Grummt I. 2002. The nucleolar remodeling complexNoRC mediates heterochromatin formation and silencing of riboso-mal gene transcription. Nat Genet 32:393–396.

Santos F, Reik W. 2003. Epigenetic marking correlates with devel-opmental potential in cloned bovine preimplantation embryos. CurrBiol 13:1116–1121.

Stopka T, Skoultchi AI. 2003. The ISWI ATPase Snf2h is required forearly mouse development. Proc Natl Acad Sci U S A 100:14097–14102.

Sudarsanam P, Winston F. 2000. The Swi/Snf family nucleosome-remodeling complexes and transcriptional control. Trends Genet16:345–351.

Sung LY, Gao S, Shen H, Yu H, Song Y, Smith SL, Chang CC, Inoue K,Kuo L, Lian J, Li A, Tian XC, Tuck DP, Weissman SM, Yang X, ChengT. 2006. Differentiated cells are more efficient than adult stem cellsfor cloning by somatic cell nuclear transfer. Nat Genet 38:1323–1328.

Thompson PMGT, Kok M, White PS, Brodeur GM. 2003. CHD5, a newmember of the chromodomain gene family, is preferentiallyexpressed in the nervous system. Oncogene 22:1002–1011.



Tsukiyama T, Wu C. 1995. Purification and properties of an ATP-dependent nucleosome remodeling factor. Cell 83:1011–1020.

Tsukiyama T, Daniel C, Tamkun J, Wu C. 1995. ISWI, a member of theSWI2/SNF2 ATPase family, encodes the 140 kDa subunit of thenucleosome remodeling factor. Cell 83:1021–1026.

Ulyanova NP, Schnitzler GR. 2005. Human SWI/SNF generatesabundant, structurally altered dinucleosomes on polynucleosomaltemplates. Mol Cell Biol 25:11156–11170.

Varga-Weisz PD, Wilm M, Bonte E, Dumas K, Mann M,Becker PB. 1997. Chromatin-remodelling factor CHRAC con-tains the ATPases ISWI and topoisomerase II. Nature 388:598–602.

Wakayama T, Tateno H, Mombaerts P, Yanagimachi R. 2000. Nucleartransfer into mouse zygotes. Nat Genet 24:108–109.

Wang F, Zhang R, Beischlag TV, Muchardt C, Yaniv M, Hankinson O.2004. Roles of Brahma and Brahma/SWI2-related gene 1 in hypoxicinduction of the erythropoietin gene. J Biol Chem 279:46733–46741.

Whitworth KM, Agca C, Kim JG, Patel RV, Springer GK, Bivens NJ,Forrester LJ, Mathialagan N, Green JA, Prather RS. 2005. Tran-scriptional profiling of pig embryogenesis by using a 15-K memberunigene set specific for pig reproductive tissues and embryos. BiolReprod 72:1437–1451.

Young LE, Sinclair KD, Wilmut I. 1998. Large offspring syndrome incattle and sheep. Rev Reprod 3:155–163.



developmental capacity of porcine nuclear transfer embryos correlate with levels of...

Documents