Aberrant Nucleo-cytoplasmic Cross-Talk Results in Donor Cell mtDNA Persistence in Cloned Embryos

Abstract

Mitochondrial DNA is an extranuclear genome normally maternally inherited through the oocyte. However, the use of nuclear transfer can result in both donor cell and recipient oocyte mitochondrial DNA persisting through to blastocyst and being transmitted to the offspring. The degree of donor mitochondrial DNA transmission appears to be random and currently no evidence exists to explain this phenomenon. To determine whether this is a dilution factor or directly related to the transcriptional status of the donor cell in respect of mitochondrial DNA transcription factors, we have generated sheep nuclear transfer embryos using donor cells: (1) possessing their full mitochondrial DNA complement, (2) those partially depleted, and (3) those depleted but containing residual levels. For each donor type, donor mitochondrial DNA persisted in some blastocysts. It is evident from the donor cells used that nuclear-encoded mitochondrial DNA transcription and replication factors persist even after mitochondrial DNA depletion, as do transcripts for some of the mitochondrial-encoded genes. These cells are therefore still programmed to drive mitochondrial DNA replication and transcription. In nuclear transfer-derived embryos, we have observed the persistence of these nuclear-encoded mitochondrial DNA transcription and replication factors but not in those embryos generated through in vitro fertilization. Consequently, nucleo-mitochondrial interaction following nuclear transfer is out of sequence as the onset of mitochondrial replication is a postimplantation event.

THE mitochondrial genome (mtDNA) is located within the inner membrane of the mitochondrion, the organelle that generates cytoplasmic ATP in all eukaryotic cells. This extranuclear genome contributes 13 of the 63 polypeptides associated with the electron transfer chain (ETC), 22 tRNAs, and two rRNAs (Anderson et al. 1981). Consequently, the ETC is the only entity in mammalian cells encoded for by two genomes, namely the nuclear and mitochondrial genomes. Transcription and replication of mtDNA is regulated by nuclear-encoded transcription and replication factors, such as mitochondrial transcription factor A (TFAM) (see Clayton 1998 for review). These factors interact with polymerase γ (PolG), an mtDNA-specific polymerase, to initiate and drive mtDNA replication (Ropp and Copeland 1996).

mtDNA is normally unimaternally inherited in a homoplasmic manner (Birky 1995, 2001). This is certainly the case for intraspecific crossing although clearly sperm mtDNA can also be transmitted to offspring resulting from interspecific crossing (Gyllensten et al. 1991; Shitara et al. 1998). Furthermore, supplementation into a recipient oocyte with donor cytoplasm or mitochondria can result in bimaternal mtDNA transmission ranging from 0 to 100% transmission (Jenuth et al. 1996; Laipis 1996; Meirelles and Smith 1997; Brenner et al. 2000). Bimaternal transmission has also been observed following nuclear transfer (NT), where both donor cell mtDNA (D-mtDNA) and recipient oocyte mtDNA (R-mtDNA) are transmitted to the resultant offspring.

However, studies related to mtDNA following NT have focused mainly on the transmission of mtDNA to the fetus and/or offspring (see St. John et al. 2004a for review). For example, the transmission of D-mtDNA neither has been detected in ovine NT-derived offspring (Evans et al. 1999) nor investigated in ovine embryos. In other species, the persistence of D-mtDNA in NT-derived embryos is variable. It has been detected in bovine embryos derived by both intraspecific NT (Steinborn et al. 1998a; Do et al. 2002) and interspecific NT (Meirelles et al. 2001) although not in all cases (Meirelles et al. 2001; Takeda et al. 2003). Additionally, D-mtDNA was detected in caprine embryos derived by interspecific NT (Jiang et al. 2004) and in embryos derived by cross-species NT (Chang et al. 2003; Yang et al. 2003). Again, the levels of D-mtDNA detected in NT-derived embryos are equally variable. Quantitative studies on bovine NT-derived embryos show levels typically between 0 and 13% (Steinborn et al. 1998a; Meirelles et al. 2001; Takeda et al. 2003) although one report suggests contributions as high as 63% (Meirelles et al. 2001).

The contribution of D-mtDNA observed in both NT-derived embryos and offspring could be related to the ratio of D-mtDNA to R-mtDNA prior to oocyte reconstruction. For example, blastomere D-mtDNA from varying ranges of development has been detected in offspring at a variance of 0 to 57% (Hiendleder et al. 1999; Meirelles et al. 2001). To this extent, Steinborn et al. (1998a,b) proposed that single blastomeres derived from a 92-cell morulae would contain less mtDNA than single blastomeres from 24-cell morulae and consequently an NT offspring derived from the former donor cell would contain less D-mtDNA. This was based on evidence that D-mtDNA levels remain constant throughout preimplantation development (Steinborn et al. 1998a) and on the assumption that D-mtDNA is equally partitioned between daughter blastomeres and that D-mtDNA is progressively diluted out at each cell division. Indeed, two bovine offspring derived from 92-cell morula blastomeres as nuclear donors contained less D-mtDNA than two offspring derived from 24-cell morula blastomeres (Steinborn et al. 1998b). Despite this, in the same study, two offspring derived from 52-cell morula blastomeres contained less D-mtDNA than the offspring derived from 92-cell morula blastomeres.

In somatic cell NT (SCNT)-derived mice, greater levels of D-mtDNA transmission were detected in offspring derived using adult fibroblasts compared to offspring derived using either immature Sertoli cells or cumulus cells as nuclear donors (Inoue et al. 2004). These investigators proposed that the variance in levels of transmission arose from the technique utilized to introduce the donor cell rather than the D-mtDNA copy number introduced into the R-oocyte, as the mitochondrial copy numbers were similar for the various donor cells (Inoue et al. 2004). However, if the D-mtDNA outcome were related to the ratio of D-mtDNA to R-mtDNA prior to oocyte reconstruction, then somatic cells having greater mtDNA copy numbers should facilitate the level of D-mtDNA persisting in embryos and subsequent transmission to the offspring.

Equally important, in a somatic donor cell, the mechanisms regulating mtDNA transcription and replication are active. A variety of transcription and replication factors mediate these interactions. Two key factors are recruited to initiate mtDNA replication, TFAM, and the mitochondrial-specific PolG. mtDNA replication is a tightly regulated process, which is initiated postimplantation. Failure to activate this cross-talk between nuclear-encoded mtDNA transcription and replication factors prior to gastrulation is lethal for embryonic development (Larsson et al. 1998). To this extent, we have previously hypothesized that a somatic cell not only would require nuclear reprogramming to regulate chromosomal gene expression but also would be vital to regulate nucleo-cytoplasmic interaction and specifically mtDNA replication (St. John et al. 2004a,b).

In this respect, we have generated NT embryos using donor cells possessing varying compositions of mtDNA to determine whether the amount of mtDNA in the donor cell indeed determines its persistence to the blastocyst stage. We also have sought to determine whether it was possible through the regulation of the number of donor mtDNA copies to restrict the blastocyst mtDNA composition to that of R-mtDNA only, thus generating homoplasmic blastocysts. We further analyzed mtDNA replication factors to determine whether their aberrant expression in donor cells prior to NT could account for the persistence of the D-mtDNA that we observed.

MATERIALS AND METHODS

All chemicals were from Sigma-Aldrich (Dorset, UK), unless otherwise stated.

Isolation of cells from primary cultures depleted of mtDNA:

Ovine (PDFF2 and SFF1) and caprine (SFF2) fetal fibroblast primary cultures were depleted of mtDNA according to the protocol of King and Attardi (1996). Briefly, cells were cultured in Dulbecco's modified Eagle's medium containing 4500 mg/liter glucose and 1 mm pyruvate and supplemented with 1 mm l-glutamine, 5% fetal bovine serum (FBS) (v/v), and 50 μg/ml uridine in the presence (depletion) or absence (nondepletion) of 50 ng/ml ethidium bromide (EthBr). Cells were collected for analysis at various time points.

Real-time PCR to determine mtDNA depletion:

Each reaction consisted of 20 ng/μl of total DNA, 10 μl absolute QPCR SYBR green mix (Abgene, Epsom, UK), 0.5 μm of ND3 F and R primers (see Table 1), ddH₂O to 20 μl. Reactions were performed in a Rotorgene-3000 real-time PCR machine (software Version 6, Corbett Research, Mortlake, NSW, Australia). A series of 10-fold dilutions of the target product from a known concentration was used for standards. Reaction conditions were 95° for 15 min and then 50 cycles of 95° for 5 sec, 60° for 20 sec, and 72° for 20 sec (extension and data acquisition on FAM/Sybr channel). Melt curve analysis was conducted by ramping from 72° to 99°, 1° per step, and data was acquired from the FAM/Sybr channel.

TABLE 1.

PCR primers

Primer	Gene specificity	Sequence (5′–3′)	Annealing temperature	Nucleotide position	GenBank accession no.
A. mtDNA D-loop and gene-specific primers
D1/F	D-loop	CTTCCCACTCCACAAGCC	15532–15550	NC001941
D2/R	D-loop	CATTAATTATATTATGCCCATGCTTACC	56°	16399–16427	NC001941
D4/R	D-loop	CGGCATGGTGAACAAGCTC	16056–16075	NC001941
β-Actin F	β-Actin	CATGTATGTGGCCATCCAGGCTG	63°	129–151	AF129289
β-Actin R	β-Actin	GACGCCGCAGTGGCCATCTC	412–431	AF129289
ND1 F	ND1	CTCAACACTAGCAGAAACAA	52°	3304–3324	NC001941
ND1 R	ND1	TTAGTTGGTCGTAACGGAAT	3580–3600	NC001941
ND3 F	ND3	CATCGCATTCTGACTTCCCCAAC	58°	9524–9547	NC001941
ND3 R	ND3	GCTATGGTGAGCATTGTGTTTAGATTAG	9712–9740	NC001941
COX I F	COX I	CCT ACT ATT TGG TGC CTG AGC	58°	5387–5408	NC001941
COX I R	COX I	AAG CCC CTG CAT GGG CTA AGT	5744–5765	NC001941

Nuclear donor	Primer	Sequence (5′–3′)	Annealing temperature	Product size (bp)
B. AS–PCR primers to identify the persistence of D-mtDNA in NT-derived embryos
PDFF2	FAM-379 R1	TGCTTATATGCATGGGGTAG	55°	202
JOE-379 R2	TGCTTATATGCATGGGGTAA
481/379 F (Common)	CCTCATGCATATAAGCACGT
SFF1	FAM-84 F1	CCAACATATCTTATGTCTGTCC	55°	148
JOE-84 F2	CCAACATATCTTATGTCTGTCT
	232-252 R (Common)	ACGTGCTTATATGCATGAGG

FAM-labeled primers are tagged at the 5′-end with GAAGGTGACCAAGTTCATGCT; JOE-labeled primers are tagged with GAAGGTCGGAGTCAACGGATT.

Proliferating cell nuclear antigen staining:

Cells were grown on coverslips and washed in PBS prior to fixation in 70% ethanol for 20 min at −20°. After fixation, the cells were washed in PBS and blocked in 5% BSA in PBS for 45 min at room temperature. Mouse anti-proliferating cell nuclear antigen (PCNA) antibody (1:2000) was incubated for 1 hr at 39°. Secondary antibody was rabbit anti-mouse FITC conjugated (Dako, Ely, UK) and was used at 1:20 for 45 min at room temperature. Coverslips were mounted in Vectashield (Vector Laboratories, Burlingame, CA) containing DAPI for nuclear counterstaining.

Oocyte recovery and in vitro maturation:

Sheep ovaries were collected from a local slaughterhouse and transported to the laboratory in PBS at 25°. Briefly, in vitro maturation of recovered oocytes was carried out by aspirating follicles (2–3 mm in diameter) and only good-quality oocytes with compact cumulus cells and evenly pigmented cytoplasm were selected. Selected oocytes were washed in HEPES-buffered TCM 199 (Gibco Life Technologies, Glasgow, UK) containing 10% FBS (Gibco Life Technologies) and cultured in maturation medium bicarbonate-buffered TCM 199 (Gibco Life Technologies) supplemented with 10% FBS, 5 μg/ml follicle stimulating hormone (Vetrepharm, Galway, Ireland), 5 μg/ml luteinizing hormone (Vetrepharm), 1 μg/ml estradiol, 0.3 mm sodium pyruvate, 100 μm cysteamine, and 50 μl/ml gentamycin. Groups of 40–45 oocytes were transferred to four-well dishes (Nunc, Roskilde, Denmark) and incubated with 500 μl of the maturation medium covered with mineral oil at 39° in an atmosphere of 5% CO₂ and maximum humidity.

At 15 hr post-onset of maturation (hpm), oocytes were exposed to HEPES-buffered synthetic oviduct fluid (H-SOF) containing 300 IU/ml of hyaluronidase and then vortexed for 4–5 min to remove cumulus cells. The denuded oocytes were incubated for 15 min in H-SOF containing 5 μg/ml of Hoechst 33342. A portion of cytoplasm containing the extruding anaphase I–telophase I spindle was removed in H-SOF plus 7.5 μg/ml cytochalasin B (CB). The aspirated karyoplast in the pipette was visualized under fluorescent light to confirm the presence of chromatin. Enucleated oocytes were cultured in maturation medium until the injection of donor cells. Nondepleted (mtDNA⁺) and partially depleted (mtDNA^PD) cultured cells and those depleted to residual levels (mtDNA^R) were used as donor cells. For each experiment, the mtDNA⁺ cells were cultured to the same time points as the mtDNA^PD and mtDNA^R cells prior to use. At 21–22 hpm, single donor cells were transferred into the perivitelline space of enucleated oocytes, and the donor–cytoplast couplets were exposed to a single electric pulse of 25 V/cm for 60 μsec in 0.3 m mannitol without calcium ions using a Multiporator (Eppendorf, Cambridge, UK). Fused couplets were placed in the incubator in synthetic oviduct fluid medium (mSOF) supplemented with 4 mg/ml BSA until activated.

Activation and in vitro culture of NT embryos:

The fused embryos were activated by initiating 5 min of incubation in 5 μg/ml calcium ionophore (A23187) followed by culture in 10 μg/ml cycloheximide and 5 μg/ml CB for 5 hr at 39°. Activated embryos were extensively washed in culture medium before transfer into 50-μl drops of mSOF supplemented with 2% (v/v) basal medium Eagle amino acids, 1% (v/v) minimum essential medium nonessential amino acids, and 4 mg/ml BSA covered with mineral oil and incubated at 39° in an atmosphere of 5% CO₂, 5% O₂, 90% N₂, and maximum humidity. Cleaved embryos were transferred into fresh culture medium containing 5% FBS on days 2 and 5 of culture.

PCR amplification and DNA sequencing:

Total DNA was isolated from the nuclear donor cell lines using the Puregene DNA isolation kit (Flowgen, Nottingham, UK), according to cell culture DNA isolation protocol. Recipient oocyte DNA was prepared by repeated freezing and thawing of the embryos, which had been supplemented with 20 μl of sterile ddH₂O, as described by Findlay et al. (1996). PCR amplification of the ovine D-loop region was performed in 50-μl reactions. Each reaction contained 200 ng of total DNA, 1× PCR buffer (BioLine, London), 1.5 mm MgCl₂ (BioLine), 200 μm dNTPs (BioLine), 0.5 μm of each relevant forward and reverse primer (see Table 1), and 2.5 units BioTaq DNA polymerase (BioLine). Reaction conditions were 95° for 5 min followed by 35 cycles of 94° for 45 sec, 55° for 30 sec, and 72° for 45 sec and then 72° for 3 min using an MJ Research (Watertown, MA) PTC-200 DNA engine (GRI, Braintree, UK). PCR products were resolved on 2% agarose gels (BioLine) at 100 V for 1 hr and product size was confirmed against a 100-bp DNA ladder (Gibco BRL, Glasgow, UK). The PCR products were excised from the agarose gels and purified for DNA sequencing using the QIAquick gel extraction kit (QIAGEN, London) as described in the manufacturer's protocol. The purified mtDNA was then sequenced according to the automated direct sequencing protocol (Hopgood et al. 1991) using a GeneAmp 9700 and the ABI PRISM BigDye terminator cycle sequencing ready reaction kit (Applied Biosystems, Foster City, CA) with both the respective forward and reverse primers. Electrophoresis of cycle sequencing products was performed on an ABI PRISM 377 sequencer (Applied Biosystems). Both primers and sequences were verified for specificity using Blast Search (http://www.ncbi.nlm.nih.gov/entrez/BLAST/).

Allele-specific real-time PCR:

Reactions were performed in 20-μl volumes where 2 ng/μl of each of the gel-purified mtDNA samples following amplification of the D-loop region was used as template to ensure that the primers amplified only mtDNA and not nuclear pseudogenes (see St. John and Schatten 2004). Eighteen microliters of master mix consisting of 1× buffer (BioLine), 1.5 mm MgCl₂ (BioLine), 200 μm dNTPs (BioLine), 0.5 μm of each of the two allele-specific (AS) primers tagged with a sequence specific to either Ampliflour SNP FAM or JOE for allelic discrimination (see Table 1), 5 μm of the common reverse primer (see Table 1), 0.5 μl of each of the 20× Ampliflour SNP FAM and JOE primers, and 2 units Biolase Diamond DNA polymerase (BioLine) was added. BioTaq Diamond DNA polymerase lacks 5′ → 3′ exonuclease activity, which ensures that only bases present in the template are amplified (St. John and Schatten 2004). The specificity of the primers to their intended targets was confirmed through mismatch assays. Real-time PCR was performed on the Rotorgene-3000 real-time PCR machine (software Version 6, Corbett Research). The reaction conditions were one cycle at 95° for 3 min followed by 50 cycles of 94° for 30 sec, annealing for 30 sec (see Table 1), and 72° for 30 sec. The samples were run in duplicate, the AS experiments were repeated three times, and the product size was verified through gel electrophoresis, as described above.

Standards were prepared by subcloning PCR products into the pCR4-TOPO vector using the TOPO TA cloning kit for sequencing (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. Inserts were amplified from within the new construct to determine their presence and sequenced to confirm their allelic specificity using M13F (5′-tgt aaa acg acg gcc agt-3′) and M13R (5′-cag gaa aca gct atg acc-3′) primers.

RT–PCR:

RNA was isolated from mtDNA⁺, mtDNA^PD, and mtDNA^R cells using the RNAqueous-4PCR kit (Ambion, Austin, TX) according to the manufacturer's instructions. A total of 800 ng/μl RNA was used as template for the generation of cDNA using the reverse transcription system kit (Promega, Madison, WI). RT was performed for 2 hr at 42° followed by 15 min at 72°. We screened for β-actin transcripts to ensure for uniformity in loading and ensured that all nuclear-encoded primers spanned at least one intron. As the mtDNA genome does not possess introns, we ensured that all RNA templates were purified from DNA contamination by performing appropriate no-enzyme controls in the RT–PCR. cDNA reactions were performed in 50-μl volumes consisting of 1× PCR buffer (Bioline), 1.5 mm MgCl₂ (Bioline), 200 μm nucleotide mix (Bioline), and 0.5 μm of each primer, and 2.5 units BioTaq DNA polymerase (Bioline). Reaction conditions were: initial denaturation at 94° for 5 min followed by 35 cycles of denaturation at 94° for 30 sec, annealing (see Table 1) for 30 sec, and extension at 72° for 45 sec with a final extension of 72° for 5 min. The products were resolved on 2% agarose gels.

Assessment of mitochondrial membrane potential:

Mitochondrial membrane potential was assessed with 5,5′,6,6′-tetrachloro-1,1′,3,3′tetraethyl-benzimidazolylcarbocyanine iodide (JC-1, Molecular Probes, Eugene, OR). Cells were incubated in 10 μg/ml JC-1 for 30 min, rinsed, and observed with the Zeiss imaging system (Zeiss, Hertfordshire, UK) that uses a Carl Zeiss AxioCam HR digital camera and software (version 5.05) on an inverted Axiovert 200 microscope and the data collected using the Axiovision software (version 3.1). Cells were excited at 485 nm and emission was detected in both red and green channels. Red fluorescence indicates active mitochondria with high membrane potential through the formation of J-aggregates. Green fluorescence is representative of JC-1 remaining in its monomeric form in less active mitochondria.

Fluorescence-activated cell sorter analysis:

Cells were stained with 5 μg/ml JC1 and analyzed using a Becton-Dickinson (San Jose, CA) fluorescence-activated cell sorter (FACS) FACS 440, with a single argon ion laser. The standard sheath fluid used was 0.22 μm of filtered autoclaved isotonic saline. Cells were first gated for their forward and side light scatters, prior to measuring JC-1 fluorescence. JC1 excitation was at 488 nm FL1-H and emission was collected at 530 nm ± 30 nm and 575 nm ± 30 nm). Cells were stained with 10 μg/ml of propidium iodide (PI). PI excitation was at 488 nm FL3-H and sort regions were selected on the basis of side vs. forward scatter and emission collected at 630 nm ± 30 nm.

Immunocytochemistry:

Antibodies were used to identify the mtDNA-encoded cytochrome c oxidase (COX) I (Molecular Probes), and PolG (Abcam, Cambridge, UK) and TFAM (Santa Cruz Biotechnologies, Santa Cruz, CA), the nuclear-encoded transcriptional and replication regulators of mtDNA. Cells were fixed in 2% formaldehyde (BDH Laboratory Supplies, Poole, Dorset, UK) for 1 hr, permeabilized with 1% (v/v) Triton X-100, and blocked with 10% (v/v) normal goat serum and 2.5% BSA in PBS. Antibody (5 μg/ml) was added and the cells were incubated at 37° for 1 hr. Cells were rinsed with 0.1% Triton X-100, labeled with appropriate Alexa Fluor 488 anti-mouse or anti-rabbit or 593 anti-rabbit or anti-goat immunoglobulin G (IgG) secondary antibodies (Molecular Probes), counterstained with DAPI (Vectashield), and viewed either by an Axioplan 2 imaging system, HBO100 (Zeiss), and captured with a AxioCam HR digital camera (Zeiss) and the data collected using the Axiovision LE Rel 4.2 program (Zeiss) or by the Leica DM IRE2 confocal microscope with Leica TCS SP2 scanner (Leica Microsystems, Buckinghamshire, UK) and analyzed using the Leica confocal software (Leica). For confocal microscopy, FITC excitation was at 488 nm and detected between 500 and 535 nm, and rhodamine excitation was at 594 nm and detected between 600 and 700 nm. For fluorescence microscopy, DAPI was excited at 395 nm and detected at 420 nm; FITC was excited at 488 nm and detected between 515 and 565 nm, and rhodamine was excited at 450 nm and detected between 580 and 590 nm. For comparative analysis of protein expression in embryos through confocal microscopy, the auto gain function was switched off and the same gain and photomultiplier settings were used for each embryo analyzed. Each channel was adjusted to remove the effects of bleed before the images were taken. FITC and rhodamine images were acquired at the same time so each pair of images corresponded to the same section.

Western blotting:

mtDNA⁺, mtDNA^PD, and mtDNA^R protein lysates were reduced prior to SDS–PAGE by the addition of 5× SDS sample buffer: 0.35 g DTT (Sigma, St. Louis), 1.56 ml 1.0 m Tris/HCl (pH 6.8), 2.5 ml glycerol (10% v/v in ddH₂O), 0.5 g SDS (Bio-Rad, Hemel Hempstead, UK), 100 μl bromophenol blue (5% w/v in ddH₂O) to 80 μg of total protein and boiled for 5 min at 100°. Protein was run on a 10% polyacrylamide gel, alongside an SDS molecular weight marker (range 27–180 kDa; Sigma), for 45 min at 200 V. Proteins were blotted onto Hybond-ECL nitrocellulose membrane (Amersham Biosciences, Little Chalfont, UK) for 1 hr at 400 mA. Membranes were blocked with 5% (w/v) dried milk powder (Safeway, UK) in PBS and slowly agitated for 1 hr at room temperature. After washing extensively in TBS-T (TBS; 1.5 m NaCl, 0.2 m Tris, and ddH₂0 to 1 liter, pH 7.6, plus 0.1% (v/v) Tween 20) for 40 min, the blot was incubated in primary antibody with one of COX I, the nDNA-encoded 30 kDa iron/sulfur (Fe/S) protein of complex I of the ETC (Molecular Probes), and PolG and TFAM for 1 hr at room temperature. After further washing in TBS-T for 40 min, the blot was incubated in the secondary antibody (rabbit anti-mouse IgG conjugated to horseradish peroxidase) for 1 hr before washing in TBS-T. Visualization of protein bands was through the ECL Western blotting system (Amersham Biosciences). Silver staining confirmed initial equal loading of protein.

Statistical analysis:

Embryo development:

Differences in the frequency of oocytes fusing successfully, fused embryos reaching the cleavage stage, and fused embryos reaching the blastocyst stage between nuclear donor groups were compared using the Fisher's exact test (Agresti 1992). P-values <0.05 were considered statistically significant.

Mean D-mtDNA levels:

Differences in mean D-mtDNA (%) levels in prematernal zygotic transition (MZT) and MZT embryos, post-MZT embryos, and all embryos among the nuclear donor groups were analyzed using Kruskall-Wallis one-way ANOVA on ranks. P-values <0.05 were considered statistically significant.

Persistence of D-mtDNA:

Differences in the frequency of pre-MZT and MZT embryos, post-MZT embryos, and all embryos that contained D-mtDNA among the nuclear donor groups were compared using the Fisher's exact test (Agresti 1992). P-values <0.05 were considered statistically significant.

RESULTS

Depletion of mtDNA from somatic cells:

To determine whether somatic cell mtDNA persists in NT-generated embryos, we have depleted two primary ovine (PDFF2; SFF1) and one caprine (SFF2) fetal fibroblast donor cell lines of their mtDNA composition to varying degrees. These cell lines were treated with the commonly used mtDNA depletion agent, ethidium bromide (King and Attardi 1996). The donor cells consisted of those possessing their full content of mtDNA (mtDNA⁺), those partially depleted (mtDNA^PD), and those prepared to almost complete depletion but possessing residual populations of mtDNA (mtDNA^R). Rates of depletion were determined by quantitative real-time PCR (Table 2). It is evident that the vast majority of depletion takes place between days 0 and 8 (PDFF2: 100–4.104%; SFF1: 100–1.146%). Although complete depletion was possible only to 0.012% in PDFF2 mtDNA^R cells, this low level of residual D-mtDNA demonstrated the sensitivity of our assay. Indeed, the small differences existing between days 106 and 119 (Table 2) indicate that our treatment was at the level from which further depletion proved impossible. Consequently, for the generation of mtDNA^PD-NT and mtDNA^R-NT reconstructed oocytes using PDFF2 donor cells, we chose cells after 58 and 109 days of EthBr treatment, respectively. For each experiment, mtDNA⁺ cells were used from the same time points in culture as the ethidium bromide-treated cells.

TABLE 2.

Real-time PCR analysis of mtDNA levels in PDFF2 and SSF1 cells treated with EthBr

		No. of days in EthBr
	Control/untreated	5	8	34	75	106	119
A. PDFF2
Mean total mtDNA (ng/μl) ±SD	13.104 ± 1.565	8.676 ± 0.847	0.538 ± 0.136	0.016 ± 0.002	0.005 ± 0.002	0.002 ± 0.001	0.002 ± 0.002
% mtDNA (control/untreated)	100.00	66.206	4.104	0.119	0.040	0.012	0.019

No. of days in EthBr
	Control/untreated	3	7	11	14	18	24	38	59
B. SSF1
Mean total mtDNA (ng/μl) ±SD	9.560 ± 1.471	0.702 ± 0.245	0.110 ± 0.037	0.008 ± 0.004	0.005 ± 0.002	0.031 ± 0.005	0.013 ± 0.003	0.000 ± 0.000	0.001 ± 0.000
% mtDNA (control/untreated)	100.000	7.342	1.146	0.086	0.052	0.324	0.137	0.002	0.014

mtDNA levels in mtDNA-depleted (A) PDFF2 and (B) SSF1 cells and their untreated controls analyzed by quantitative real-time PCR analysis. The mean concentration (in nanograms per microliter) of mtDNA was determined by comparing the amount of mtDNA-encoded ND3 gene product generated in each sample to that generated by a series of standards of known concentration. mtDNA levels in mtDNA-depleted PDFF2 and SSF1 cells are also presented as a proportion of mtDNA levels in untreated PDFF2 or SSF1 cells, taken from the same time points in culture, where untreated cell mtDNA levels are represented as 100%.

It is nevertheless apparent that these donor cells are almost completely devoid of expressible mtDNA-encoded protein after 11 days (Figure 1E) of EthBr treatment and this persists through to no expression after 22 days (Figure 1F), as determined by immunocytochemistry (ICC). Western blotting demonstrates the loss of mtDNA-encoded protein expression in these cells to very low levels with the continued suppression in those cells subsequently used for NT reconstruction (Figure 1G).

Figure 1.

The loss of mtDNA-encoded protein during mtDNA depletion of donor cells. (A–F) Antibody staining of the mtDNA-encoded COX I on PDFF2 cells untreated (A–C) or treated (D–F) with 50 ng/μl EthBr after 11 (A and D), 16 (B and E), and 22 (C and F) days in culture. It is evident that most of the mtDNA-encoded protein is lost very early on in the treated populations with only residual amounts retained. This is indicative of the low levels of mtDNA still persisting at these stages of EthBr treatment. Nuclei were counterstained with DAPI (stained blue). (G) Western blot analysis of the mtDNA-encoded COX I and the nuclear-encoded 30 kDa Fe/S protein of complex I (loading control). Protein lysates were extracted from untreated and EthBr-treated PDFF2 and SFF1 cells and resolved by SDS–PAGE. The presence of the 30-kDa protein demonstrates that nuclear-encoded proteins for the ETC are still being recruited although mtDNA-encoded protein is detectable only at very low levels in any of the treated cell populations. Res, residual mtDNA population.

However, the mitochondria of cells having undergone the depletion process are functional as evidenced by the persistence of the mitochondrial membrane potential through staining with JC-1. JC-1 is a dual stain, which can identify high membrane potential through J-aggregates (red fluorescence) and low membrane potential through J-monomers (green fluorescence; Smiley et al. 1991; Cossarizza et al. 1996). It is evident that the mitochondria of the treated cells are morphologically very different. Imaging for the J-aggregates in both treated and untreated cells revealed a very different morphology between the two cell types. The long tubal structures observed in the untreated cells are replaced by roundish spheres in the depleted populations of cells (Figure 2A). Both treated and untreated populations of SFF2 cells were synchronized to have similar cell cycle profiles prior to FACS analysis for both JC-1 and PI. Both treated and untreated cells exhibit mitochondria with high membrane potential, suggesting that the mitochondria in the treated cells are still functioning (Figure 2B). However, the mitochondria of the treated cells have a greater proportion of mitochondria with low membrane potential in comparison with the untreated cells, suggesting that metabolic function has most likely switched from oxidative phosphorylation (OXPHOS) to glycolysis. This is reflected in the requirement of these cells to be supported in culture by uridine. The data for PI sorting verify the cell cycle synchronicity of the cells and reveal that neither the treated nor the untreated populations are apoptotic or necrotic (data not shown).

Figure 2.

Morphology and activity of mitochondria in cell-cycle-phase-synchronized untreated and EthBr-treated SFF2 cells stained with the mitochondrial-specific fluoroprobe JC-1. (A) Morphology of untreated and EthBr-treated live SFF2 cells through J-aggregates (red fluorescence) shows the long tubular structures in the untreated cells being replaced by roundish structures in the EthBr-treated cells. (Bi–Biii) Untreated SSF2 cells contain proportionately more mitochondria with high membrane potential than low membrane potential, indicated by the large RPE (red) fluorescent peak (Bii) relative to the FITC (green) fluorescent peak (Biii). (Biv–Bvi) Treated SSF2 cells contain a similar proportion of mitochondria with high and low membrane potential, indicated by fluorescent peaks of similar intensity in the red (Bv) and green (Bvi) channels, respectively.

Embryo development:

PDFF2 donor cells were assessed to determine their synchronicity in G₀ of the cell cycle (Campbell et al. 1996) and PCNA analysis indicated that only a low proportion of the cells in both groups (control 8%; depleted 5.2%) were proliferating under these culture conditions (data not shown). Using these PDFF2 cells, we generated 183 mtDNA⁺, 95 mtDNA^PD, and 70 mtDNA^R reconstructed embryos from 206, 130, and 75 oocytes, respectively (see Table 3). There was a significant difference between cleaved oocytes for the mtDNA⁺ group against the mtDNA^R (P < 0.03) and mtDNA^PD groups (P < 0.03). However, there was no significant difference in rates of development between any of the groups from fusion to the blastocyst stage (P > 0.05).

TABLE 3.

Developmental competence of ovine NT embryos reconstructed using EthBr-treated and untreated PDFF2 donor cells

Donor cells	No. of oocytes	No. of fused embryos	No. of cleaved embryos	No. of NT blastocysts	No. of cells/blastocyst (mean ± SEM)
mtDNA+	206	183 (88.3%)a	166 (90.7%)a	61 (33.3%)a	60.7 ± 8.6a
mtDNAR	75	70 (93.3%)a	56 (80.0%)b	15 (21.4%)a	68.5 ± 7.8a
mtDNAPD	130	95 (73.1%)b	77 (81.1%)b	30 (31.2%)a	ND

mtDNA⁺ cells were cultured to the same time point as the mtDNA^R and mtDNA^PD populations. Superscripts a and b denote a significant difference between nuclear donor cell types in the efficiency of fusion (mtDNA^PD vs. mtDNA⁺, P < 0.0003, and mtDNA^PD vs. mtDNA^R, P < 0.0004) and fused embryos reaching the cleavage stage (P < 0.03). There were no significant differences for fused embryos reaching the blastocyst stage for any of the nuclear donor cell types (P > 0.05). ND, not determined.

mtDNA analysis:

The persistence of D- and R-mtDNA in 43 ovine NT embryos derived from mtDNA⁺, mtDNA^PD, and mtDNA^R PDFF2 nuclear donor cells was analyzed using AS-PCR (see Table 4). Analysis of mixtures of known concentrations of D-mtDNA and R-mtDNA determined the sensitivity of our AS-PCR assay to be sufficient to detect <0.1% of D-mtDNA for each AS-PCR primer pair. In total, 18 of 20 mtDNA⁺-, 9 of 9 mtDNA^PD-, and 5 of 14 mtDNA^R-derived embryos contained D-mtDNA. The persistence (presence of heteroplasmy) of D-mtDNA at all stages of embryo development was significantly different between the mtDNA⁺ vs. mtDNA^R (P < 0.002) and mtDNA^PD vs. mtDNA^R (P < 0.003) groups (see Table 4), with no significant difference between the mtDNA⁺ and mtDNA^PD groups. Analysis between premajor MZT (1-cell to 16-cell stage; Memili and First 1999) embryos revealed that the persistence of D-mtDNA in mtDNA⁺ embryos was not significantly different from mtDNA^R or mtDNA^PD embryos (P > 0.05) but was significantly different between mtDNA^R and mtDNA^PD embryos (P < 0.03). However, in the post MZT group (32-cell, morulae and blastocysts), the persistence of D-mtDNA in the mtDNA⁺ embryos was significantly different from those embryos generated from mtDNA^R cells (P < 0.02), while there was no significant difference between the mtDNA^PD and mtDNA^R embryos (P > 0.05).

TABLE 4.

Analysis of the persistence of donor mtDNA and the degree of heteroplasmy in PDFF2 NT-generated embryos

Donor cell	2-cell embryo	4-cell embryo	8-cell embryo	16-cell embryo	32-cell embryo	Blastocyst	Hatched blastocyst	Mean ± SD
mtDNA+	0.05	0.06	0.00	0.02	0.13	0.03 (n = 2)	0.00	0.90 ± 2.02
0.05	0.02
0.10	0.52
0.45	0.92
0.61 (n = 2)	1.00
0.93
3.66
8.72
mtDNAPD	0.86	0.30	0.27	0.84	1.68	0.08	ND	0.57 ± 0.49
0.48
0.23
0.39
mtDNAR	0.00 (n = 2)a	0.00	ND	0.00	ND	0.00 (n = 4)	0.00	0.01 ± 0.02
0.02	0.01
0.02
0.04
							0.08

ND, not determined.

^aThese were a 1-cell and a 3-cell embryo.

Analysis of the overall mean D-mtDNA levels revealed significantly greater levels in mtDNA⁺-derived embryos compared to mtDNA^R embryos (P < 0.0007; see Table 4) and in the mtDNA^PD embryos compared to mtDNA^R embryos (P < 0.0003). However, there was no significant difference between mtDNA⁺ and mtDNA^PD embryos (P > 0.05). In the premajor MZT group, there was no significant difference between mtDNA⁺ embryos and those reconstructed using mtDNA^PD and mtDNA^R cells (P > 0.05), although a significant difference was observed between mtDNA^PD and mtDNA^R embryos (P < 0.02). In the post-MZT mtDNA^R embryos (see Table 4), there was significantly less donor mtDNA present than in the mtDNA⁺ (P < 0.003) and mtDNA^PD (P < 0.03) embryos. Although the mean D-mtDNA level in the mtDNA⁺ embryos was greater than in the mtDNA^PD post-MZT embryos, the difference was not significant (P > 0.05). Interestingly, mean D-mtDNA levels were greater in the post-MZT embryos compared to the premajor MZT-stage embryos in the mtDNA⁺ group (P = 0.05) but not in the mtDNA^PD group (P > 0.05) or in the mtDNA^R (P > 0.05) embryos.

To determine whether our findings were cell line specific or the persistence of D-mtDNA was eliminated following NT-oocyte reconstruction, we analyzed another ovine fetal fibroblast cell line (SFF1). Interestingly though, this cell line took a considerably shorter time to deplete and these cells were considered to be appropriate for mtDNA^R-NT after 38 days in culture (see Table 2). Again, as with the PDFF2 experiments, SFF1 mtDNA⁺ cells were taken from the same time point in culture as the mtDNA^R series. We observed similar fusion, cleavage and blastocyst rates of development between the mtDNA⁺ and mtDNA^R donor cells (see Table 5). The cleavage rates were similar to the PDFF2 cell line, but the frequency of fused oocytes developing to blastocyst were significantly higher for the PDFF2 mtDNA⁺ group (P < 0.003). Interestingly, however, the number of blastomeres per blastocyst was significantly greater for the mtDNA^R group (see Table 5; P < 0.05), unlike those analyzed for the PDFF2 line (see Table 3).

TABLE 5.

Developmental competence of ovine NT embryos reconstructed using EthBr-treated and untreated SFF1 donor cells

Groups	No. of oocytes	No. of fused couplets (%)	No. of cleaved embryos (%)	No. of blastocysts (%)	No. of cells/blastocyst (mean ± SEM)
mtDNA+	106	94 (88.7)a	81 (86.2)a	15 (16.0)a	48.7 ± 4.5a
mtDNAR	106	86 (81.1)a	65 (75.6)a	16 (18.6)a	69.8 ± 6.2b

mtDNA⁺ and mtDNA^R donor cells were cultured to the same time points. Superscripts a and b within a column denote significant differences. The number of cells per blastocyst was significantly greater in the mtDNA^R donor cell group compared to the mtDNA⁺ donor cell group (P < 0.05). No significant differences were observed between the mtDNA⁺ and mtDNA^R donor cell groups for the efficiency of fusion, cleavage, and blastocyst outcome (P > 0.05).

mtDNA analysis of SFF1 mtDNA⁺ (n = 6) and mtDNA^R (n = 8) embryos indicated that there were no significant differences in the persistence of D-mtDNA between mtDNA⁺ and mtDNA^R embryos overall or between premajor MZT and post-MZT groups (P > 0.05). Although the mean mtDNA levels were higher in mtDNA⁺ compared to mtDNA^R embryos, the differences were not significant (P > 0.05). Similarly, premajor MZT mtDNA⁺ embryos contained higher D-mtDNA levels than mtDNA^R embryos, but again the differences were not significant (P > 0.05). Furthermore, there was no significant difference between the two sets of embryos at the post-MZT stage of embryonic development.

Comparison of the embryos generated using the two donor cell lines (PDFF2 and SFF1) revealed similar patterns of persistence and levels of D-mtDNA between mtDNA⁺ embryos and mtDNA^R embryos (P > 0.05).

Transcription factor analysis:

To determine why extremely low levels of mtDNA may persist following mtDNA depletion to 0.002% and 0.119%, as with the SFF1 and PDFF2 cell lines, respectively, we analyzed the expression of the mtDNA-specific nuclear-encoded PolG and TFAM. TFAM is involved in both mtDNA replication, in combination with PolG, and transcription (Clayton 1998; Kajuni 2004). We note that in these primary cell lines PolG and TFAM expression still persists and, in the case of TFAM, seems upregulated throughout the depletion process (see Figure 3). This suggests that the cell would still be capable of mtDNA transcription and replication through the persistence of such nucleo-cytoplasmic interaction. To this extent, mtDNA-encoded transcripts show gradual and variable rates of decline for the PDFF2 line, exemplified by the loss of COX I (day 34), but persistent low levels of transcripts for COX II, ND 1, and ND 3 even at day 34 of EthBr treatment when the level of mtDNA is 0.119% (see Figure 4) and thereafter (data not shown). However, ICC analysis of in vitro fertilization (IVF)- and NT-generated embryos shows that the maternal genome differentially regulates PolG and TFAM. At the two-cell stage, TFAM and PolG are still detectable in both IVF- and NT-generated embryos (data not shown). In comparison, the NT-generated embryos still express these factors at the four-cell stage while the IVF embryos have diminished levels (Figure 5), indicating that only the maternal genome has downregulated the requirement for mtDNA replication.

Figure 3.

The persistence of PolG and TFAM in donor cells prior to NT. (A) Detection of PolG and TFAM by immunocytochemistry. Donor cells were labeled with primary antibodies to PolG and TFAM as described in materials and methods. The relative amount of each protein during the mtDNA depletion process was also monitored by Western blotting. The proteins were detected in both cell lines used as nuclear donors for NT (PDFF2 and SSF1) at their described molecular weights of ±140 kDa (PolG, B) and ±25 kDa (TFAM, C), in control (Con) and cells with residual amounts (Res), as well as in PDFF2 cells depleted for 11 and 22 days (11d and 22d, respectively). A positive control (Pos) is also shown. Throughout the depletion process both mtDNA⁺ and mtDNA^R cells express PolG and TFAM, indicating that these cells are still actively capable of transcribing and replicating the mitochondrial genome and that mtDNA depletion does not reduce the expression of the these transcriptional and replication factors. Indeed, PolG is present at similar levels in both cell lines and seems reduced only in PDFF2 with residual mtDNA. Interestingly, the production of TFAM seems upregulated following the mtDNA depletion process in both cell lines.

Figure 4.

RT–PCR analysis of mtDNA-encoded transcripts throughout the depletion process. Nuclear encoded β-actin transcript levels are unaffected by the mtDNA depletion of PDFF2 cells using EthBr. In contrast, mtDNA-encoded ND1 and ND3 and COX I and II transcript levels are sequentially reduced after 5 to 14 days of EthBr treatment. MtDNA-encoded transcripts are either eliminated (COX I) or continually persist at low levels (ND1, ND3, and COX II) after 34 days of EthBr treatment. mtDNA-encoded transcript levels are expressed at similar levels in untreated cells grown in parallel with the EthBr-treated cells.

Figure 5.

Confocal microscopy imaging of COX I, PolG, and TFAM labeling in IVF- and NT-generated embryos. IVF and NT embryos from the four-cell stage were labeled with COX I (control) and PolG and TFAM to determine the regulation of their presence throughout very early preimplantation embryonic development. Whole embryos are shown through confocal microscopy (×63 objective and ×3 digital zoom). It is evident that COX I protein persists to contribute to ETC-generated ATP. However, the two vital transcription and replication factors, TFAM and PolG, are present at very diminished levels in the IVF-generated embryos at the four-cell stage but persist in the NT-derived embryos. COX I was detected using the Alexa Fluor 488 anti-mouse, TFAM using the Alexa Fluor 594 anti-goat, and PolG with the Alexa Fluor 488 anti-rabbit IgG secondary antibodies (Molecular Probes). Controls using only the secondary antibodies showed no staining.

DISCUSSION

We have shown that near-complete mtDNA-depleted somatic donor cells can be effectively used to generate NT embryos that develop to the blastocyst stage. These are capable of progressing to the hatched blastocyst stage, indicating their readiness for implantation. Although it is generally accepted that ethidium bromide is a mutagenic agent, at low concentrations it can specifically target the mitochondrial genome (King and Attardi 1996). It has been extensively used to generate mtDNA-depleted cells for the production of cell hybrids to study mtDNA-type disease associated with mutations within the mtDNA genome and to study nuclear–mitochondrial cross-talk (Joseph et al. 2004). Our prolonged culture of donor cells in ethidium bromide did not suggest an adverse affect on either set of donor cells. Although the percentage of blastocysts generated using the PDFF2 cell line was higher for the mtDNA⁺ embryo population, this was not significantly different from the other populations, indicating that there were no adverse effects on preimplantation development through depletion. Indeed, for the SFF1 population, the blastocyst rates were even more marginal. Furthermore, the near-complete elimination of D-mtDNA resulted in a significantly greater number of blastomeres per blastocyst. In this instance, this result suggests an increased developmental competence for embryos generated from D-mtDNA^R cells. Despite the increase in the number of cells, these embryos have a lower total cell number compared to our in vitro-produced ovine embryos (Maalouf et al. 2005) and those reported by other groups (Gardner et al. 1994; Papadopoulos et al. 2001).

For the two donor cell lines that we utilized, it is clear that the vast difference in mtDNA content between the two cell types affects the rates of depletion (see Table 2). This likely reflects the differing ATP requirements of these two cell lines derived from mixed populations of ovine fibroblasts. Indeed, different cell types possess very different numbers of mitochondria. For example, neuronal and muscle cells are high ATP users (reviewed in Moyes et al. 1998; discussed in St. John et al. 2004a,b). In addition, the number of mtDNA copies per mitochondrion also reflects a cell's bias to ATP generation. Again, neurones and muscle cells possess between 8 and 10 mtDNA copies per mitochondrion and favor the OXPHOS pathway, the cell's major generator of ATP (reviewed in Moyes et al. 1998). Furthermore, there are significant differences in mtDNA copy number between, for example, skeletal and cardiac muscle (Miller et al. 2003) and peripheral blood mononuclear cells and subcutaneous fat (Gahan et al. 2001). However, for the generation of homoplasmic SCNT embryos, the choice of donor cell may undoubtedly affect the level of D-mtDNA transmitted.

Somatic cells depleted of their mtDNA content are unable to survive in culture unless supported by anaerobic metabolic supplementation, as were our D-mtDNA^R cells during the depletion process and prior to NT, and thus utilize the far less productive ATP-generating glycolytic pathway (King and Attardi 1996). The loss of mtDNA molecules is clearly reflected in the decrease in mtDNA-encoded protein associated with the ETC (see Figure 1). However, the loss of ETC-encoded protein does not affect mitochondrial function as determined by JC-1 activity, although it is clear that mitochondrial morphology is disturbed, as reported in other studies following mtDNA depletion (Gilkerson et al. 2000). In many respects, this switch from aerobic to anaerobic metabolism for our donor cells matches those enhanced metabolic conditions that are recognized as promoting SCNT efficiency and the resultant murine NT embryonic development (Chung et al. 2002).

The use of more developmentally advanced blastomeres to regulate the amount of donor mtDNA persisting in the embryo and transmitted to the offspring has been demonstrated to an extent (Steinborn et al. 1998b). However, our data clearly show that even the residual D-mtDNA present in our mtDNA^R cells can persist to the blastocyst stage in 5 of the 10 PDFF2 embryos that we analyzed. Consequently, to ensure that homoplasmic embryos are generated, D-mtDNA must be completely eliminated.

This persistence and potential transmission of D-mtDNA to offspring may reflect the localization of the D-mtDNA in its clustering around the donor nucleus, as indicated in mtDNA supplementation experiments where the site of D-mtDNA introduction can affect transmission (Meirelles and Smith 1998). It could also reflect the donor nucleus's desire to preferentially select for its own mtDNA at the expense of R-mtDNA, as demonstrated in primate–human cybrids (Moraes et al. 1999). Furthermore, the persistence of a very low population of mtDNA and low levels of mtDNA-encoded transcripts and protein would suggest that an active DNA replication mechanism may exist to perpetuate such an outcome. Recently, a short burst of mtDNA replication in the pre-MZT phase in mouse IVF embryos has been reported (McConnell and Petrie 2004), which could account for the persistence or, at least, the preservation of the D-mtDNA detected. Although it has been hypothesized that D-mtDNA might be diluted out (Steinborn et al. 1998a,b), we have previously reported the persistence of sperm D-mtDNA following blastomere NT in rhesus macaque (St. John and Schatten 2004). This clearly indicates that if a minimal amount of supplemental mtDNA is present, then it is available for replication and transmission. Indeed, this phenomenon is not just limited to NT, but has been observed following, for example, interspecific murine mating (Gyllensten et al. 1991) and associated with human mtDNA disease (Schwartz and Vissing 2002). Consequently, any mtDNA introduced into an oocyte, postprimordial germ-cell stage can be transmitted as it would not be regulated by the oogenic “genetic bottleneck” hypothesized to function during early embryogenesis (Marchington et al. 1997) and sperm mtDNA consequently would have evaded the ubiquitin elimination process (Sutovsky et al. 1999).

Reprogramming the donor nuclear genome is also vital for the regulation of nucleo-cytoplasmic cross-talk. The continued expression of PolG and TFAM observed in our donor cells prior to NT suggests that these donor (mtDNA⁺, mtDNA^PD, and mtDNA^R) cells are still programmed to drive mtDNA replication. This is in stark contrast to ovine IVF embryos, which at the four-cell stage have very little PolG or TFAM expression. However, both these replication factors still persist in the four-cell NT-derived embryos. TFAM expression is regulated by the binding of nuclear respiratory factor 1 (NRF-1; Huo and Scarpulla 2001) to the specific binding site of the TFAM promoter (Choi et al. 2004). NRF-1 mediates mtDNA transcription in the developing murine embryo from the two-cell stage. While it is evident that disruption to NRF-1 expression leads to reduced expression of TFAM and the subsequent decrease in mtDNA copy number (Virbasius and Scarpulla 1994), in vitro studies have shown that methylation of the NRF-1 binding sites could silence the TFAM promoter (Choi et al. 2004).

Genomewide demethylation is an epigenetic DNA modification characteristic of early embryos and is partially, but abnormally, recapitulated in clones (Dean et al. 2003; Bourc'his et al. 2001). Absence of nuclear localization of DNA methyltransferases prior to the blastocyst stage precludes any active methylation of the DNA (Ratnam et al. 2002), suggesting that active genes in the donor cell, such as TFAM, will not be repressed after SCNT. Continuous expression of TFAM will lead to early replication of the mitochondrial genome in cloned embryos. This is further enhanced by the continued expression of PolG, as both replication factors are required to drive mtDNA replication (see Kajuni 2004). This will increase the probability of replication of D-mtDNA, which is in the vicinity of the donor nucleus in early cleavage stages (Sato et al. 2005). Consequently, TFAM and PolG may also be vital markers of reprogramming, since tight regulation of their expression will determine the onset of nuclear–mitochondrial cross-talk in preimplantation embryos and would therefore account for the continued persistence of the D-mtDNA even when residual levels persist in the reconstructed embryo.

Failure to regulate these coordinated nuclear–cytoplasmic events, i.e., nuclear–mitochondrial cross-talk through incomplete nuclear reprogramming, could also result in the donor nucleus driving mtDNA replication in a manner similar to its pre-NT status, for example, that of a fibroblast (St. John et al. 2004a,b). As a result, blastomeres might express mitochondrial phenotypes similarly to the donor nucleus as might subsequent differentiated cells in fetal and postnatal tissue (St. John et al. 2004a). This could have profound implications for mtDNA differentiation in both cloned offspring and NT-derived embryonic stem cells. Consequently, complete elimination of D-mtDNA and somatic cell reprogramming are molecular events that together could drive harmonic nuclear–mitochondrial cross-talk post-NT.