SEARCH

SEARCH BY CITATION

Keywords:

  • Cloning;
  • Evolutionary distance;
  • Heteroplasmy;
  • Mitochondrial DNA;
  • Replication;
  • Transcription

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Following fertilization, mitochondrial DNA is inherited from the oocyte and transmitted homoplasmically. However, following nuclear transfer, mitochondrial DNA can be transmitted from both the donor cell and recipient oocyte, resulting in a state of heteroplasmy. To determine whether the genetic diversity between donor cell and recipient cytoplast mitochondrial DNA influences development, we generated bovine embryos by fusing a donor cell to one or more enucleated cytoplasts. Analysis of mitochondrial DNA from embryos, fetal tissues, and blood samples from offspring revealed that early preimplantation embryos from two or three cytoplasts had significantly more mitochondrial DNA variants than fetal tissues. Phylogenic analysis of embryos generated using single cytoplasts divided the mitochondrial DNA sequence variants into three separate groups with various amounts of genetic divergence from the donor cell line. In heteroplasmic tissue and blood samples, the predominant mitochondrial DNA population was significantly more divergent from the donor cell than the less frequent allele. Furthermore, analysis of the mitochondrially encoded cytochrome B gene showed that two heteroplasmic alleles encoded for different amino acids, and the ratios of mitochondrial DNA/mRNA for each allele differed significantly between tissues. The degree of evolutionary distance between the donor cell and the cytoplast and the variability in heteroplasmy between tissues may have an impact on more divergent intergeneric nuclear transfer and the use of this approach for the generation of embryonic stem cells.

Disclosure of potential conflicts of interest is found at the end of this article.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Mitochondrial DNA (mtDNA) is a 16.6-kilobase (kb) genome located in the mitochondrion of all eukaryotic cells [1]. It encodes 13 subunits of the electron transfer chain (ETC), the intracellular apparatus that generates the majority of a cell's ATP through the process of oxidative phosphorylation (OXPHOS). Although many of the subunits of the ETC are encoded by chromosomal DNA, mutation or deletion of an mtDNA gene can either lead to severe loss of cellular, tissue, and organ function or can be lethal [2].

Following fertilization of an oocyte, mtDNA is inherited strictly through the oocyte [3], as sperm mtDNA is normally eliminated by the process of ubiquitination [4], except in cases of interspecific crossing [5]. mtDNA in the mature oocyte is derived from a few copies present in 10–20 mitochondria [6], which are segregated to primordial germ cells during an early restriction event just after gastrulation and subsequently clonally expanded [7]. Consequently, these copies of mtDNA are usually identical (homoplasmic).

Somatic cell nuclear transfer (SCNT) involves the introduction of a somatic donor cell or nucleus into an enucleated oocyte [8]. The activation of this reconstructed oocyte gives rise to embryos and intra- and interspecific offspring, which have genetically identical chromosomes but not mtDNA as, in most cases, the majority of mtDNA will be derived from the recipient oocyte [9, 10]. However, mtDNA accompanying the donor cell can persist in embryos, fetuses, and offspring (reviewed in [11]). This can constitute up to 59% of their total mtDNA content [12] and thus gives rise to heteroplasmy. Such an outcome probably arises as the donor cell's mtDNA replication factors are still expressed during the early stages of preimplantation development, unlike in vitro-fertilized (IVF) embryos [13, 14]. This could result in the donor cell selectively favoring the replication of its own mtDNA. Furthermore, cloning protocols for larger mammals often use in vitro-matured (IVM) metaphase II (MII) oocytes harvested randomly from ovaries. The variation in mtDNA sequences between recipient oocytes and donor cells for the encoding genes, as observed between porcine or bovine breeds, could affect amino acid composition and thus compromise ETC function [15, 16].

Although intergeneric cloned embryos have been generated [17, [18], [19]20], they have not survived to term, perhaps because of the genetic divergence that exists between the donor cell and its recipient oocyte. Analysis of mtDNA genetic divergence between the mtDNA haplotypes of the two fusion partners has demonstrated a threshold for donor cell and recipient oocyte compatibility that intra- and interspecific (Ovis aries-O. aries) fusions appear to comply with but not intergeneric (Capra hircus-O. aries) nuclear transfer (NT) fusions [14]. This outcome is critical, as the donor cell's mtDNA genes will be compatible with its nuclear-encoded genes for functional ATP output but not necessarily with the recipient cytoplast's mtDNA genes. Such outcomes would have an impact on the potential to derive embryonic stem cells (ESCs) through intergeneric NT. It could also affect the ability of any fully differentiated cell to function efficiently, especially those cells highly dependent on OXPHOS (i.e., muscle and neuronal cells and cardiomyocytes).

Handmade cloning (HMC) most likely exacerbates the outcomes of mtDNA transmission, as the donor cell can be fused to one or more cytoplasts [21, 22], with relatively similar quantities of mtDNA from various oocyte sources being incorporated. Consequently, this provides an excellent model to study (a) the relationship between mtDNA haplotype and development under the control of a somatic nucleus, and (b) the transmission and segregation of mtDNA in embryos. Furthermore, the use of multiple cytoplasts could result in direct competition between two or more mtDNA haplotypes. If one of these haplotypes is more compatible (i.e., closer in terms of genetic distance) with the mtDNA haplotype of the donor cell, then it may be preferentially replicated, resulting in the other haplotypes not being propagated. However, if the evolutionary distances between the populations of mtDNA are all within appropriate ranges for the donor cell nucleus, then these mtDNA haplotypes will be replicated with the same efficiency, resulting in a stable state of heteroplasmy.

To understand the effects of mtDNA genetic diversity on SCNT outcome, we have analyzed embryos from the zygote to the blastocyst stage that were generated with a somatic cell and one cytoplast. We have extended this to embryos generated with two and three cytoplasts to determine whether a genetically closer population of mtDNA has a selective advantage. Analysis of multiple mtDNA populations in fetuses and offspring allowed us to determine whether mtDNA from different backgrounds undergoes selection or segregation postimplantation and whether a mixed population is capable of coexisting in live offspring. We have also determined, using tissues from an aborted fetus, whether there is selection for one allele over another at the transcriptional level.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Handmade Cloning

Bovine embryos were generated by HMC using one, two, or three cytoplasts. In addition, tissue samples from 10 fetuses that spontaneously aborted between day 200 and day 280 and blood samples from six live offspring (>6 months) [22, 23] were analyzed. All embryos, fetuses, and offspring were generated from the donor cell line Flavor, except for animal 748, which was generated with Pandora. Pathological analysis of the spontaneously aborted fetuses detected a range of conditions from normal (n = 3) to severe (n = 4) heart/liver/kidney/muscle abnormalities. The remaining fetuses (n = 3) showed a range of minor defects in one or two organs as described above.

The cumulus cells and zona pellucida were removed from IVM MII oocytes at 18–22 hours postmaturation (hpm), bisected and reconstructed with a donor cell using electrofusion, followed by parthenogenetic activation (PA) at 21–34 hpm and in vitro cultured according to previously described methods [22]. A more detailed description is provided in the supplemental online Materials and Methods.

Polymerase Chain Reaction Amplification and Sequencing of the mtDNA D-Loop

Total DNA was extracted from individual embryos and stored frozen in 2-μl sterile double-distilled H2O (ddH2O), using the single freeze-thaw protocol, as described in [13]. For one-cytoplast embryos, a minimum of five embryos were included from each stage of preimplantation development. For analyses, embryos were divided into 1-cell (n = 6), 2–6-cell (n = 25), 8–16-cell (n = 10), and blastocyst (n = 9) stages. The two- and three-cytoplast embryos were also analyzed at the 1-cell (n = 16), 2–6-cell (n = 22), 8-cell to morula (n = 12), and blastocyst (n = 10) stages. DNA was extracted from donor cells, fetal tissue (n = 10), and blood samples (n = 6) using the Puregene DNA Isolation Kit (Qiagen Ltd., West Sussex, U.K., http://www1.qiagen.com), according to the manufacturer's protocols [14]. Two hundred nanograms of DNA from tissues and cells or 2 μl of DNA from embryos was amplified in 50-μl reactions containing 1× polymerase chain reaction (PCR) buffer, 1.5 mM MgCl2, 200 mM dNTPs (all from Bioline, London, http://www.bioline.com), 0.5 μM CO1F (CACCATCAACCCCCAAAGCT), and 0.5 μM CO3R (TTGGGTTAAGCTACATCAAC) to produce 974 base pairs (bp) of the mtDNA D-loop. The reaction was performed at 35 cycles in an MJ Research PTC-200 thermocycler (GRI, Braintree, U.K., http://www.gri.co.uk), with an initial denaturation of 95°C for 5 minutes followed by 35 cycles of 94°C for 45 seconds, 54°C for 30 seconds, and 72°C for 2 minutes. When only low levels of product were obtained, 2 μl of PCR product was amplified with CO1F and CO2R (CCTGAAGAAAGAACCAGATG) (annealing temperature, 55°C) to generate 498 bp for DNA sequencing.

All PCR products were resolved on a 2% agarose gel at 100 V for 1 hour and purified using the QIAquick Gel Extraction Kit (Qiagen). Purified PCR products were then sequenced according to the automated direct sequencing protocol [24] using a GeneAmp 9700 and the ABI PRISM BigDye terminator cycle sequencing-ready reaction kit (Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com) with the appropriate forward and reverse primers. Electrophoresis of cycle sequencing products was performed on an ABI PRISM 377 sequencer (Applied Biosystems).

Subcloning of PCR Products

Purified PCR product was ligated into a pCR4-TOPO cloning vector (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) and cultured, as described previously [25], to identify sequences of individual alleles within a heteroplasmic population.

Sequence Alignment and Haplotype Analysis

The first and last 20 bp of each sequence were removed so that 457 bp were analyzed (nucleotides 15,807–16,264 of the Bos taurus mtDNA reference sequence; GenBank accession no. V00654; [26]). For embryos constructed with one cytoplast, singletons were removed and replaced with the consensus sequence, as described previously [14], to identify the most frequent variants between the different mtDNA lineages. All sequences from single-cytoplast embryos were aligned using the ClustalW program with a Phylip output format [27] and then analyzed using the phylogenetic programs of PHYLIP, version 3.6 [28]. Pairwise evolutionary distances were calculated using DNADIST, which used the F84 model of nucleotide substitution (PHYLIP [28]). Phylogenetic trees were constructed from sequence alignments, first bootstrapping sequences 1,000 times in SEQBOOT, and then by maximum likelihood analysis using the DNAML and DRAWGRAM programs (PHYLIP [28]). Sequences generated from two- and three-cytoplast embryos and tissue and blood samples contained various numbers of heteroplasmic sites. Following subcloning, individual alleles were identified, and those containing the higher peaks on the chromatogram, representative of the overall sample, were designated the “major” alleles; those containing the lower peaks were designated the “minor” alleles. Sequences for major and minor alleles were aligned and analyzed as for the single-cytoplast embryos.

RNA Isolation

RNA was extracted from cells using the DNA-free RNA Isolation Kit for RT-PCR (Ambion, Austin, TX, http://www.ambion.com) according to the manufacturer's protocol for tissue samples. Contaminating DNA was eliminated by adding 2 U of DNase1 (Ambion) and incubating at 37°C for 1 hour.

Reverse Transcription

cDNA was produced using the Promega Reverse Transcription System (Promega, Southampton, U.K., http://www.promega.com). Each 20-μl reverse transcription (RT) reaction contained 800 ng of RNA, 5 mM MgCl2, 1× RT buffer, 1 mM dNTPs, 12.5 μg ml−1 oligonucleotide primers, 1 unit μl−1 RNasin, 0.75 unit μl−1 AMV-RT, and sterile ddH2O. In addition, no-enzyme and no-RNA controls were performed. The reaction was performed at 42°C for 2 hours in an MJ Research PTC-200 thermocycler. Prior to Allele Specific-PCR (AS-PCR), 100 μl of ddH2O was added to the cDNA.

Allele Specific-PCR Analysis of Tissue Samples

DNA and cDNA were amplified using forward (CCTCTGTTACCCATATCTGCCG) and reverse (CTAGAATTAGTAAGAGGGCCCC) primers producing 521 bp (nucleotides 14,704–15,225; GenBank accession no. V00654) of the Cytochrome B (CytB) gene. Initial denaturation at 95°C for 5 minutes was followed by 40 cycles of 94°C for 30 seconds, 59°C for 30 seconds, and 72°C for 50 seconds and a final extension at 72°C for 8 minutes. PCR products were subcloned and sequenced using M13 forward (TGTAAAACGACGGCCAGT) and reverse (CAGGAAACAGCTATGACC) primers, as described previously [14].

Each 20-μl reaction contained 2 μl of PCR product or purified plasmid, 1 × buffer (ABgene, Epsom, U.K., http://www.abgene.com), 1.5 mM MgCl2 (ABgene), 800 μM dNTPs (Chemicon International, Temecula, CA, USA), 0.5 μM FAM-tagged (GAAGGT-GACCAAGTTCATGCTATTACGGGTCTTACACTTTTC) or JOE-tagged (GAAGGTCGGAGTCAACGGATTATTACGGGTCTTACACTTTTA) AS-primer, 5 μM common reverse primer (ATGTATGGGATTGCTGATAAG), 0.5 μl of each of the 20 × Amplifluor SNPs FAM and JOE primers (Chemicon), and 1 U of Thermostart DNA Polymerase (ABgene). Reagents were pipetted using a CAS-1200 liquid handling system (Corbett Life Science, Sydney, New South Wales, Australia, www.corbettlifescience.com). AS-PCR was performed using the Rotorgene-3000 real-time PCR machine and software (Corbett Life Science). The reaction consisted of 8 cycles of 95°C for 120 seconds, 54°C for 15 seconds, and 72°C for 15 seconds, followed by 50 cycles of 95°C for 20 seconds, 54°C for 15 seconds, and 72°C for 15 seconds. Detection was through the FAM and JOE channels during the annealing phase. Samples were amplified alongside two sets of 10-fold dilution standards, one for each allele.

Statistical Analysis

All statistical analysis was performed using GraphPad Prism version 4.0 (GraphPad Software, Inc., San Diego, http://www.graphpad.com). All data sets were first tested for Gaussian distribution using the Kolmogorov-Smirnov test. The proportional data for cell fusion and in vitro development of embryos were analyzed by the χ2 test, whereas the differences in the number of major alleles from offspring and fetal samples were clustered into one of the three groups of the phylogenic tree and then analyzed by the χ2 test. For the phylogenic groups, contingency tables were constructed on the basis of the assumption that each group contained the same number of major alleles. The Kruskal-Wallis test and then Dunn's multiple comparison test were used to determine the significance of the following: the phylogenic distribution of single-cytoplast embryos; evolutionary distance and stage of development (one-cell; two- to six-cell; eight-cell to morula; and blastocyst, fetus, and offspring) for one, two, and three cytoplasts and for major and minor alleles in heteroplasmic embryos, fetuses, and blood samples; the percentage of heteroplasmic sites in two- and three-cytoplast embryos, fetal tissue, and blood samples; and overall differences between all developmental stages. For relative proportions of alleles 1 and 2 for major and minor alleles, six replicates were converted to logarithmic values. Two-way analysis of variance (ANOVA) was used to test for significance between the values for DNA and mRNA for different tissue types, and one-way ANOVA was used to test for global differences between all six DNA and mRNA samples and for differences in transcriptional ratios among tissues. Bonferroni post hoc tests were used to test between individual tissue types. In all cases, p < .05 was deemed significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Handmade Cloning

Fusion rates were significantly higher in HMC embryos constructed with two and three cytoplasts than those with one cytoplast (p < .0001; Table 1). Overall, day 7 blastocyst rates were significantly higher for the PA group compared with all HMC embryos (Table 1; p < .05). However, HMC embryos generated with two or three cytoplasts resulted in significantly higher blastocyst rates than those from one cytoplast (p < .0001). Furthermore, embryos reconstructed with a single cytoplast and cultured individually produced poor-quality blastocysts.

Table Table 1.. Fusion efficiencies and blastocyst rates for embryos derived from PA and HMC embryos reconstructed with 50% (one cytoplast), 100% (two cytoplasts), or 150% (three cytoplasts) cytoplasmic volume
Thumbnail image of

mtDNA D-Loop Sequence Analysis of Single-Cytoplast Embryos

To determine whether mtDNA genetic divergence between the donor cell and the recipient oocyte would influence preimplantation development in embryos generated with one cytoplast, we analyzed the hypervariable region 2 of the D-loop from 50 embryos. Calculation of pairwise evolutionary distances between each embryo and the donor cell line revealed three clusters (Fig. 1). Group 1 contained 10 embryos, each with very similar mtDNA sequences. Group 2 consisted of 19 embryos, which clustered with the donor cell line and contained several subgroups. Group 3 contained 21 embryos with three separate subgroups. The evolutionary distances were significantly different among the three groups identified in Figure 1 (p < .0001). The mean evolutionary distances (± SEM) for embryos in group 1 (0.071% ± 0.0012%) and group 3 (0.07% ± 0.0008%) were significantly greater than that for group 2 (0.011% ± 0.0014%; p < .001 for both), which contained the donor cell line.

thumbnail image

Figure Figure 1.. Phylogenetic tree for one-cytoplast embryos calculated from sequence data using a maximum likelihood model. Data were bootstrapped 1,000 times using SEQBOOT to ensure the validity of the tree. Embryos are labeled according to their developmental stage. Numbers preceding the developmental stage are sample identification numbers. Abbreviation: blast, blastocyst.

Download figure to PowerPoint

Overall comparison of developmental stage and evolutionary distance revealed a significant difference (p < .02; supplemental online Fig. 1). This reflected the lower mean distance from the donor cell line (0.0074% ± 0.00235%) for the 1-cell embryos, which was significantly lower than that of the 2–6-cell-stage embryos (0.056% ± 0.005%; p < .05), 8–16-cell stage embryos (0.049% ± 0.0089%; p < .05), and the blastocyst (0.05135% ± 0.011%; p < .05). At all preimplantation developmental stages other than the one-cell stage, embryos were generated to a maximum evolutionary distance of 0.078374% from the donor cell line.

Heteroplasmy in Two- and Three-Cytoplast Embryos

D-loop sequences from 60 preimplantation embryos generated using two (n = 30) and three (n = 30) cytoplasts were analyzed for heteroplasmic mtDNA variants. Whereas 54 embryos exhibited various numbers of heteroplasmic sites, ranging from 2 (0.44% of total numbers of base pairs per sequence) to 32 (7%), the remaining 6 were homoplasmic. There was no significant variability in the occurrence of heteroplasmy between two- and three-cytoplast embryos at various stages of preimplantation development (one-cell, two- to six-cell, eight-cell to morula, and blastocyst) or between stages of preimplantation development.

Heteroplasmy in Tissue and Blood Samples

To determine the degree of heteroplasmic variants that could coexist throughout postimplantation development and in offspring, we analyzed liver samples from 10 cloned fetuses that had spontaneously aborted and blood samples from six live offspring generated using a range of cytoplasts (1–6). Tissue samples from seven fetuses possessed only one predominant mtDNA population, whereas three were heteroplasmic, harboring one of 2, 4, and 30 variant sites. Heteroplasmy was observed in three of six blood samples each containing two mtDNA populations that varied at 1, 16 and 17 sites. Comparison of the number of heteroplasmic sites throughout development demonstrated a significant overall reduction (p < .02; Fig. 2), with fetal liver samples (0.776% ± 0.65%) having significantly fewer variants than the one-cell (2.476% ± 0.62%; p < .05) and the two- to six-cell (3.225% ± 0.604%; p < .05) embryos.

thumbnail image

Figure Figure 2.. Percentage (±SEM) of heteroplasmic variants identified in a 457-base pair region of the mitochondrial DNA D-loop in preimplantation embryos generated from two or three cytoplasts, fetal tissue, and blood from live offspring generated using between two and six cytoplasts. Overall, the percentage of variants was significantly different throughout development (p < .05), as determined by the Kruskal-Wallis test. Significance between individual stages was determined using Dunn's multiple comparisons tests; ∗ denotes p < .05 significant difference from fetal tissue.

Download figure to PowerPoint

Evolutionary Distance from the Donor Line

To determine whether the predominance of one mtDNA population over another was related to its evolutionary distance from the donor cell line, the genetic distances between the donor cell line and the major and minor alleles for all samples were analyzed. The maximum evolutionary distance from the donor cell was 0.0787% (Fig. 3). Although the mean values were not significantly different between major and minor alleles at each stage of development (p > .05), the major alleles had a greater mean evolutionary distance from the donor cell line than the minor alleles in two- to six-cell embryos, eight-cell to morula-stage embryos, fetal tissue, and live offspring (Fig. 3). There was, however, an overall significant difference between the individual stages of development for the major alleles (p < .05), specifically between the one-cell embryos (0.058% ± 0.005%) and fetal tissue (0.0724% ± 0.0014%; p < .05) and blastocyst stage (0.048% ± 0.0094%) and fetal tissue (p < .01). Combining both major and minor alleles for each developmental stage resulted in the overall evolutionary distance also being significantly different throughout developmental (p = .0211), with just the blastocyst stage (0.051% ± 0.006%) and fetal tissue (0.0685% ± 0.004%) being significantly different from each other (p < .05).

thumbnail image

Figure Figure 3.. Evolutionary distance (mean ± SEM) between mitochondrial DNA alleles in embryos and tissues and the donor cell line, as determined by pairwise analysis of sequences. There were no significant differences between two- and three-cytoplast embryos at any stage of development, determined using the Kruskal-Wallis test. The combination of both major and minor alleles resulted in the overall evolutionary distance being significantly different throughout development (p = .0211), as determined by the Kruskal-Wallis test. Differences between stages of development were determined by Dunn's multiple comparison tests, where a next to two stages of development refers to their significance to each other (a, p < .05). There was also an overall significant difference between the individual stages of development for the major alleles (p < .05). Individual comparisons are denoted (b, p < .05; c, p < .01).

Download figure to PowerPoint

Maximum Likelihood Analysis of Major and Minor Alleles in Comparison with the Donor Cell Lines

To determine whether the selection of mtDNA populations within heteroplasmic clones was dependent upon their phylogenetic relationship to the donor cell line, sequences were analyzed using maximum likelihood. These included all fetal and offspring samples and a selection of two- and three-cytoplast embryos. For heteroplasmic embryos and samples, both major and minor alleles were included. Except for two outliers (Fig. 4), the sequences clustered to the three large groups described for the single-cytoplast embryos. The major alleles for the fetal and offspring samples clustered to group 1 only, with both group 1 and group 3 having significantly increased evolutionary distance compared with group 2 (p < .01). Only one minor fetal allele clustered to group 2, whereas two other minor offspring alleles clustered to the outliers (Fig. 4). One offspring (blood 748), which was homoplasmic, clustered to group 2, containing the donor cell line Flavor. However, this was generated using the donor cell line Pandora, which clustered to group 1. This suggests that a slightly increased evolutionary distance, within a viable range, is beneficial for development. This divergence is similar to that between Bos taurus and B. indicus (0.066%–0.076%), from which live offspring have been successfully generated using both embryonic cells, isolated from a preimplantation embryo [29], and a somatic cell [15].

thumbnail image

Figure Figure 4.. Phylogenetic representation of evolutionary relationships between donor cell lines and major and minor alleles from a representative sample of two- and three-cytoplast embryos and all fetal and offspring samples. The tree was calculated from mitochondrial DNA sequence data using the maximum likelihood model after the data had been bootstrapped 1,000 times. Where two populations were detected, major and minor alleles are labeled. Numbers preceding the developmental stage (e.g., 317, 62) are the sample identification numbers. Embryos are labeled according to their developmental stage. All samples were generated using Flavor as the donor cell line, except that blood sample 748 was generated using Pandora. Abbreviations: blast, blastocyst stage; mor, morula stage.

Download figure to PowerPoint

The D-loop sequence for Flavor shared the greatest level of homology with the Angus breed (GenBank accession no. DQ520592). In addition, those single-cytoplast embryos clustered in a subgroup with the donor cell in group 2 (Fig. 1) also corresponded with the B. taurus Angus sequence. Other embryos within group 2 shared the greatest level of homology with the B. taurus Heinan Yellow breed (GenBank accession no. DQ887764). Embryos in group 1 were determined to be of the B. indicus Brahman breed (GenBank accession no. DQ166132), or the B. taurus New Zealand Jersey breed (GenBank accession no. AF389178), whereas those in group 3 corresponded to the B. taurus Holstein (GenBank accession no. AY998841) or New Zealand Jersey breeds, which shared greater homology with the B. indicus Brahman breed.

Distribution of Heteroplasmic Variants Among Tissues

To determine whether major and minor alleles would be uniformly segregated during fetal development, we analyzed heart, liver, lung, kidney, muscle, and spleen tissues from a single aborted fetus generated using six cytoplasts. DNA sequencing of the D-loop determined that each of the tissues possessed two mtDNA populations. These alleles differed at four sites and thus had almost identical evolutionary distances from the donor cell line (0.06849% vs. 0.06852%). Furthermore, both clustered to group 1, but in different subgroups. The proportion of alleles 1 and 2 for each tissue were determined by analysis of the CytB gene using AS-PCR (supplemental online Fig. 2) and ranged from 50.7% to 65.0% (p > .05; Fig. 5). However, these proportions were not equivalent for mRNA. For both heart and liver, the percentage of allele 1 transcribed was lower than for their respective DNA levels (heart: mRNA, 35.0%; DNA, 62.3%; liver: mRNA, 43.3%; DNA, 53.3%; p > .05; Fig. 5). The proportion of allele 1 transcribed for kidney, muscle, lung, and spleen was higher than for their respective DNA values (Fig. 5). However, this was significantly different only in lung (p < .05). The difference among all mRNA samples was found to be significant (p < .0001). This was reflected in the low levels of allele 1 observed in the heart and liver samples, both of which were significantly lower than the values in the lung and spleen samples (p < .01).

thumbnail image

Figure Figure 5.. Cumulative frequency bar chart demonstrating the relative proportions of alleles 1 and 2 found in DNA and mRNA from six tissue samples. The black and white sections of the bar represent the percentages of allele 1 and 2, respectively. Significance between the DNA and mRNA in the different tissue types was determined using two-way analysis of variance (ANOVA). ∗, p < .05, where ∗ is shown above the bar for RNA. All mRNA samples were also significant; ∗∗, p < .0001, as determined using one-way ANOVA.

Download figure to PowerPoint

Overall, the mRNA levels for allele 1 expressed as a ratio of their respective mtDNA levels in each tissue were significantly different (one-way ANOVA; p < .001; Table 2). The heart (0.56) and liver (0.79) samples had lower levels of allele 1 for mRNA than DNA. The ratios for kidney (1.05) and muscle (1.10) were close to 1, indicating similar levels of allele 1 for mRNA and DNA. The lung (1.68) and spleen (1.51) samples both contained twice the proportion of allele 1 for mRNA compared with DNA. The ratio in heart was significantly lower compared with lung (p < .001), muscle (p < .05), and spleen (p < .001). The ratio in lung was significantly greater than that in liver (p < .001) and kidney (p < .01), whereas the ratio in spleen was also significantly greater than for liver (p < .01). These allelic variations led to changes in amino acid composition, where an A[RIGHTWARDS ARROW]G transition at nucleotide 200 of the CytB gene resulted in Asn[RIGHTWARDS ARROW]Ser, C[RIGHTWARDS ARROW]G at nucleotide 221 resulted in Thr[RIGHTWARDS ARROW]Ser, and C[RIGHTWARDS ARROW]T at nucleotide 323 resulted in Thr[RIGHTWARDS ARROW]Ile.

Table Table 2.. Ratios of allele 1 in mRNA compared with levels in DNA samples of six different tissue types
Thumbnail image of

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Although there is currently no direct evidence to associate mtDNA heteroplasmy with abnormalities and diseases arising from SCNT, mtDNA is transmitted in a manner that is dissimilar to natural fertilization (i.e., not from an oocyte that is related to the maternal set of chromosomes). Transmission tends to be a combination of donor cell and recipient oocyte mtDNA or solely from the recipient oocyte, with reports of donor cell mtDNA contributing up to 59% of the total mtDNA content of the live offspring (reviewed in [11]). Such transmission appears to be independent of the initial mtDNA content present in the donor cell at the time of oocyte reconstruction. Even donor cells depleted of the vast majority of their mtDNA content can still transmit mtDNA through to the blastocyst stage [13].

Unlike data from SCNT, our data suggest that HMC allows more than one population of oocyte mtDNA to be transmitted. Similar outcomes were observed following serial NT, where the donor cell is first introduced into an enucleated recipient oocyte and then transferred to an enucleated zygote [30]. As a result, both sources of oocyte mtDNA, but not donor cell mtDNA, were transmitted to the offspring [16]. Pronuclei transfer, where mtDNA close to the karyoplast is transferred [31], and cytoplasmic transfer (CT), where mtDNA is introduced directly into the oocyte [32, 33], also result in two populations of oocyte mtDNA being inherited. However, as with donor cell mtDNA, levels of donated or supplemented mtDNA in the offspring are variable, with segregation into one or more tissues [31].

Embryos generated with multiple cytoplasts provide an excellent model for the study of mtDNA transmission between two or more alleles in a competitive environment. This is true especially as donor cell mtDNA replication factors are still expressed prior to and postfusion through to the blastocyst, when mtDNA replication in IVF embryos is first initiated [14]. Consequently, HMC fusions allow the donor cell to select for one or both alleles from the one-cell stage onward, and these alleles will not necessarily be lost due to the large-scale degradation of mtDNA, which has been associated with bovine [34] and porcine [35] preimplantation embryos. The persistence of only one oocyte mtDNA population at various stages of development can be explained by either the randomly selected cytoplasts having arisen from the same maternal source or the preferential replication of one allele. Those mitochondria contributing to the reconstructed oocyte will have escaped the hypothetical mitochondrial bottleneck [7], a genetic filtering process presumed to take place at the primordial follicle stage, just after gastrulation, which regulates the transmission of mtDNA to the subsequent generation [36]. However, a further selective mechanism appears to take place for somatic lineages, as fetal tissues had significantly fewer heteroplasmic variants than in the one-cell and two- to six-cell embryos (Fig. 2).

The limited but necessary mtDNA divergence between the donor cell and the recipient oocyte in intra- and interspecific crosses might relate to the inheritance of the nuclear genes that drive mtDNA transcription and replication. Prior to fertilization, the diploid nucleus discards one set of chromosomes at meiosis I. However, following natural fertilization, the recombination of haploid maternal and paternal nuclei gives rise to a combination of chromosomal alleles that will encode the mtDNA-specific transcription and replication factors. Consequently, for the paternal-derived alleles encoding mtDNA transcription and replication factors, the oocyte mtDNA will be sufficiently diverse to allow for a degree of genetic divergence that could promote the faithful transcription and replication of mtDNA. Following SCNT, both alleles will have been previously conditioned to the mtDNA population of the somatic donor cell but not that of the recipient oocyte, which could compensate for the lack of chromosomal genetic diversity.

The clustering of the major alleles from heteroplasmic embryos, blood, and tissue samples to groups 1 and 3 suggests that the breeds associated with these groups may have a developmental advantage (Fig. 4). For example, Angus preimplantation embryos are more sensitive to increased temperature than Brahman embryos, and their development is significantly reduced [37]. However, it is evident from crosses between more distantly related species that increased evolutionary distance becomes detrimental. For example, the crossing of B. gaurus donor cells with enucleated B. taurus oocytes, representing an evolutionary distance of 0.108% (1.4 times the distance of the HMC offspring), resulted in 18% of reconstructed oocytes developing to the blastocyst stage, with 12% of these developing to fetuses [38]. All but one of these fetuses aborted, including one after 200 days of gestation, because of abnormal placentae [38]. More recently, cloning of the even more divergent B. javanicus [39] with B. taurus or B. indicus oocytes resulted in two pregnancies, which were lost between 30 and 90 days of gestation [40]. Although the introduction of an O. musimon donor cell into an O. aries oocyte resulted in live offspring [41], where the evolutionary distance was within a similar range [42], an O. aries-C. hircus intergeneric model resulted in preimplantation development not surpassing the 20-cell stage [14]. In this instance, the mean evolutionary distance between the donor cell and the recipient oocyte was 0.4114%.

From the above, it is evident that intra- and interspecific embryos could be generated to the blastocyst stage, which could give rise to ESCs. In addition, intergeneric ESCs have been derived following the transfer of a human donor cell into a rabbit cytoplast [43]. However, it remains to be determined whether the degree of 1 mtDNA genetic divergence influences the ability of such ESCs to function efficiently. In the mouse, various time points rotated to ESC differentiation can equate to specific stages of fetal development [44]. If a similar scenario were applied to larger mammalian ESC differentiation, then it is likely that intra- and interspecific and intergeneric ESCs would survive in their undifferentiated state and the early and mid-phases of differentiation because of the high glycolytic/anaerobic environment of the ESC culture media [45]. This would be representative of the support that the fetus would receive in utero from the mother as mtDNA mass accumulates in readiness for increased ATP output postparturition [31]. However, in their fully differentiated state, these ESCs might be compromised, just as some SCNT-derived fetuses and offspring are [38, 40].

From an mtDNA perspective, the rate-limiting factors are likely to be the incompatibility of (a) the nuclear-encoded mtDNA transcription and replication factors to mediate efficient transcription and replication, (b) mtDNA heteroplasmic populations, and (c) the nuclear- and mtDNA-encoded components of the ETC. For any SCNT-derived ESCs, the vast majority of their mtDNA content is likely to originate from the recipient oocyte source, although some donor cell mtDNA is likely to be transmitted. Previous investigations have determined that the incorporation of both donor cell and recipient oocyte mtDNA or two sources of oocyte mtDNA can result in divergent populations of mtDNA coexisting together. The segregation of these molecules can be random, as observed in porcine [46], bovine [12], and murine [47] SCNT offspring. However, such heteroplasmy can potentially encode for differing amino acid compositions because of sequence variants within various subspecies [15, 16], and this will be further exaggerated in intergeneric ESCs.

Our data show that, in heteroplasmic tissues, both molecules are indeed transcribed, but at ratios that do not directly correspond to the mtDNA ratios for the two alleles (Table 2). Furthermore, this appears to be specific to certain tissues, such as lung and heart. One explanation for this outcome might be the greater necessity for homoplasmy at the mRNA rather than the mtDNA level, so that functional ETCs could be formed to ensure maximal ATP generation. This would certainly be true for heart, liver, and lung tissue, which are high ATP users, whereas heteroplasmy might be more easily tolerated by lower ATP users. It is likely that this adaptive approach may be suitable to intra- and interspecific NT ESCs, where compatibility between the nuclear and mtDNA subunits of the ETC could be sufficient to ensure cellular function of any differentiated cell.

However, as inter-generic ESCs are generated with genetically diverse donor nuclei and recipient cytoplasts, they may not be able to produce sufficient levels of ATP to support high ATP-demanding cell types once differentiated and then challenged to function in an in vivo/aerobic environment. These incompatibilities are highlighted by somatic cell-somatic cell fusions (cybrids), where transcription and replication are not always affected but ETC activity is [48, 49], as cybrid formation and growth rates [50]. Consequently, as it is not possible to predict whether donor cell mtDNA would persist at sufficient levels for transcription of appropriate levels of species-specific mtRNA, it may be necessary to modulate the mtDNA composition of the recipient oocyte or the resultant ESCs so that more diverse forms of genetic diversity can be avoided but viable ESCs can be derived.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

We have shown through HMC that SCNT requires a certain mtDNA genetic divergence between the donor cell and its recipient oocyte for successful preimplantation, fetal development, and generation of offspring. However, this genetic divergence is restricted, with increased distance leading to embryonic or fetal failure. Such outcomes will compromise intergeneric SCNT, which has been proposed as a means of generating human ESCs, unless careful selection for appropriate recipient oocytes is undertaken. This would ensure compatibility between the nuclear- and mtDNA-encoded proteins of the ETC and their ability to generate ATP efficiently.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

E.J.B. and R.T.T. contributed equally to this work. E.J.B. is currently affiliated with The MRC Human Immunology Unit, The Weatherall Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Headington, Oxford, U.K.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information
FilenameFormatSizeDescription
SC-07-0747_Supplemental_Data.pdf24KSupplemental Data
SC-07-0747_Supplemental_Figure_1.pdf671KSupplemental Figure 1
SC-07-0747_Supplemental_Figure_2.pdf808KSupplemental Figure 2

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.