Mitochondrial DNA transmission and confounding mitochondrial influences in cloned cattle and pigs

Abstract

Although somatic cell nuclear transfer (SCNT) is a powerful tool for production of cloned animals, SCNT embryos generally have low developmental competency and many abnormalities. The interaction between the donor nucleus and the enucleated ooplasm plays an important role in early embryonic development, but the underlying mechanisms that negatively impact developmental competency remain unclear. Mitochondria have a broad range of critical functions in cellular energy supply, cell signaling, and programmed cell death; thus, affect embryonic and fetal development. This review focuses on mitochondrial considerations influencing SCNT techniques in farm animals. Donor somatic cell mitochondrial DNA (mtDNA) can be transmitted through what has been considered a “bottleneck” in mitochondrial genetics via the SCNT maternal lineage. This indicates that donor somatic cell mitochondria have a role in the reconstructed cytoplasm. However, foreign somatic cell mitochondria may affect the early development of SCNT embryos. Nuclear–mitochondrial interactions in interspecies/intergeneric SCNT (iSCNT) result in severe problems. A major biological selective pressure exists against survival of exogenous mtDNA in iSCNT. Yet, mtDNA differences in SCNT animals did not reflect transfer of proteomic components following proteomic analysis. Further study of nuclear–cytoplasmic interactions is needed to illuminate key developmental characteristics of SCNT animals associated with mitochondrial biology.

Introduction

Somatic cell cloning by nuclear transfer (SCNT) has potential agricultural applications for replicating food animals with desired genetic traits, animal transgenesis, or in conservation of endangered species. Although SCNT is a powerful tool for cloned animal production, SCNT generally results in low developmental competency and a host of abnormalities in a majority of embryos and their subsequent offspring. Pre‐ and post‐natal development is often compromised, and a variable proportion of SCNT offspring show aberrant developmental patterns and increased pre‐ and perinatal mortality [1]. Aberrant reprogramming of donor somatic cell nuclei may result in many severe problems in animal cloning [2, 3]. The interaction between the donor nucleus and the recipient cytoplast plays an important role in the efficiency of SCNT and may influence a number of important biological functions in SCNT during nuclear reprogramming. Mitochondria have a broad range of critical functions in cellular energy supply, cell signaling and programmed cell death; thus affecting embryonic and fetal development. Mitochondrial DNA (mtDNA) molecules carry genes that are essential for energy production through oxidative phosphorylation and electron transfer. This review examines mitochondrial biology and dysfunction following SCNT techniques in farm animals.

Transmission of mitochondrial DNA in cloned cattle and pigs

Although the primary purpose of cloning technology is to produce progeny genetically identical to the donor cell, mtDNA becomes particularly relevant to cloning because animals inherit most or all of their mitochondria through maternal transmission (i.e., via the recipient oocyte) [4]. Therefore, if SCNT embryos are reconstructed with heterogeneous populations of oocytes harboring different mtDNA genotypes, the genotypes of the resulting SCNT animals will differ in context of the mitochondrial genome. Porcine mtDNA sequences vary between Asian and European breeds, while less intra‐breed is detected [5, 6]. In contrast, bovine mtDNA show many intra‐breed substitutions and a host of sequence differences among maternal lineages [7, 8]. Recipient oocytes are usually generated from other cows’ ovaries; for example, from ovaries collected from slaughterhouses. Although most bovine SCNT offspring show mtDNA genotypes that differ from those of the targeted (founder) animals (Fig. 1) [9, 10], nuclear transfer is regarded as the most effective technique to exchange mtDNA in a single generation [4, 11]. The main reason for mitochondrial respiration defects in mitochondrial diseases is accumulation of pathogenic mutant mtDNAs; therefore, nuclear transfer has been considered the most effective treatment strategy to preferentially eliminate mutant mtDNA molecules or to dilute the molecules to below a threshold proportion for the expression of disease phenotypes. As expected, nuclear transplanted mito‐miceΔ showed no clinical phenotypes over the course of their lives [11]. However, the problem still remains that nuclear transfer may introduce defective mtDNA along with the nucleus into oocytes, resulting in mitochondrial heteroplasmy. Varying patterns of mtDNA transmission have been observed in SCNT animals [12, 13]. Most SCNT clones appear to contain no or very few mitochondrial haplotypes derived from SCNT donor cells [9, 14, 15, 16, 17]. However, a small fraction of clones appear to favor the replication of donor cell–derived mtDNA (D‐mtDNA) [9, 13, 18, 19]. Such variations in reported mtDNA composition could be related to differences in the SCNT procedure or differences in nucleo‐cytoplasmic interactions [12]. During normal fertilization, the oocyte contributes all of the mitochondria to the developing embryo, and the sperm mitochondria are destroyed shortly after fertilization [20]. Destruction of sperm mitochondria appears to have evolutionary and developmental advantages because the paternal mitochondria and their DNA may be compromised by the deleterious action of reactive oxygen species (ROS) encountered by sperm during spermatogenesis, storage, migration, and fertilization [21, 22]. The dramatic reduction of mtDNA content is in favor of active destruction rather than a reduced turnover of mtDNA molecules [23]. Recruitment of autophagosomal markers in the vicinity of ubiquitinated mitochondria suggests that this autophagy could also exist in mammals and be involved in the degradation of sperm‐inherited mitochondria [24]. The destruction of mtDNA content in bovine embryogenesis may be compatible with the bottleneck hypothesis proposed to explain the variable transmission of heteroplasmy in SCNT calves [23].

Figure 1.

Mitochondrial genotypes of SCNT cattle (CA1–4) produced from the same donor cell line. Nuclear transfer was performed using donor cells prepared from cumulus cells (Japanese black) and enucleated oocytes collected from a pool of ovaries (undefined, Holstein or Japanese black) [90, 91]. MtDNA genotypes were determined by PCR‐mediated single‐strand conformation polymorphism analysis using the variable region of D‐loop [6, 7]. CA3 and CA4 exhibited a mixture of genotypes with D‐mtDNA [79]

Whether D‐mtDNA in SCNT cows can be transmitted to G_ns (subsequent generations) was unknown. Mitochondria pass through a genetic bottleneck during transmission in the female germline. The rapid segregation of heteroplasmic bovine mtDNA was observed by following transmission of a single heteroplasmic mtDNA mutation within one or two generations in Holstein cattle [25, 26]. Whenever a new mutation persists and there is coexistence of two or more sequence variants in a cell, a genetic bottleneck in the female germ line will generally favor the transmission of a single mitochondrial population (homoplasmy) [13, 27, 28]. However, in SCNT animals both stable and unstable transmission of D‐mtDNA was found in porcine and bovine G_ns [10, 17]. In our study, D‐mtDNA was present at ranges of 0.1–1.0 % of total mtDNA in SCNT pigs derived from Chinese pig (Meishan breed) fibroblasts microinjected into European pig oocytes (maternal Landrace breed) [29]. However, one of the 25 G₁ progeny of the SCNT founder pigs showed an extremely high distribution of D‐mtDNA, while others did not show heteroplasmy (Fig. 2). The proportions of D‐mtDNA to total mtDNA varied among tissues; 44 % in liver, 25 % in heart, 12 % in skeletal muscle, 6.0 % in lung, 4.3 % in kidney, 0.1–1.0 % in blood, and less than 0.1 % in spleen, ovary, uterine, and skin tissue. Of the 18 (89 %) G₂ progeny of the aforementioned heteroplasmic G₁ line 16 exhibited heteroplasmy, with a number harboring D‐mtDNA (average ± SD = 12.9 ± 8.3 % in liver tissue). Similarly, some of G₁ cattle exhibited high percentages of D‐mtDNA populations (range 17–51 %), while other G₁ cattle had low or undetectable levels of D‐mtDNA. These results indicate that proportion of G₁ offspring can maintain heteroplasmy with a much higher percentage of D‐mtDNA than their SCNT dams, which may also reflect segregation distortion caused by the proposed mitochondrial bottleneck (Fig. 3). Recently, it was demonstrated that the mtDNA bottleneck is not due to a drastic decline in mtDNA copy number in early oogenesis, but rather to a small effective number of segregation units for mtDNA [30, 31], or to the genetic bottleneck occurring during post‐natal folliculogenesis [32]. It is still unclear exactly when and how the numerical restriction occurs in the maternal germ‐line, and whether the sizes of the units are different among animal species.

Figure 2.

MtDNA transmission in nuclear transfer derived pigs and their progeny [14]. D‐mtDNA (Meishan) was present in ranges between 0.1 and 1.0 % of total mtDNA in SCNT pigs (SCNT1–4). Only one of seven G₁ progeny of SCNT3 showed D‐mtDNA heteroplasmy and 16 of 18 G₂ progeny exhibited D‐mtDNA heteroplasmy

Figure 3.

Models for the transmission of mtDNA in a SCNT cow maternal lineage. D‐mtDNA (black circle) introduced into SCNT embryos could survive and be transmitted to their offspring. Random mitochondrial genetic segregation and the putative bottleneck resulted in varying patterns of mtDNA heteroplasmy in the following generation (G₁ cows)

Influence of existence of foreign somatic cell mitochondria during early development

In nuclear transfer protocols employed to date, donor cell cytoplasm, with varying proportions of viable mitochondria, is routinely transferred into recipient oocytes, with mitochondrial heteroplasmy detected post‐transfer. The physiological and morphological status of mitochondria, as well as the copy number of mtDNA molecules, differs among somatic cells, embryonic cells, and oocytes within the same animal [33, 34, 35]. Unlike mature mitochondria found in somatic cells, oocyte mitochondria are rounded or oval in appearance with a dense matrix and few arched cristae [35, 36]. Similar to sperm mitochondria, long‐term incubation in vitro, or under conditions of serum starvation, standardizing the cell cycle stage for experimental purposes may cause stress and mitochondrial breakdown in the cytoplasm. The ability to introduce exogenous mitochondria into ova was initially used to study mitochondrial dynamics and mitochondrial heteroplasmy in vivo [37]. Therefore, we investigated whether injection of somatic cytoplasm or mitochondria influenced parthenogenetic development; these results illustrate the developmental consequences of current nuclear transfer protocols [38, 39]. The exogenous mitochondria originating from donor cells directly affected parthenogenetic development. Serum starvation of donor cells prior to nuclear transfer also influenced the morphology of mitochondria and parthenogenetic development [39, 40]. These results indicate that the cytoplasm accompanying the introduction of somatic nuclei has the capacity to affect reconstructed embryo development, and that mitochondrial transfer is implicated in this deficiency. Unlike cytoplasmic or mitochondrial injections, ooplasm injection did not affect embryo development [38]. Ooplasm‐derived mitochondria were similar in morphology and more closely synchronized with those of recipient mitochondria; therefore, they may be more easily accommodated following transfer. In addition to nuclear transfer studies, ooplasmic transfer has been used as a relatively new assisted reproduction technique to supplement putatively defective cytoplasm of oocytes from patients with embryonic developmental failure [41, 42]. Such techniques may illustrate the mechanisms underlying the dynamics of persistence of foreign mitochondria and maintenance of heteroplasmy in various experimental protocols. However, it is quite difficult to detect a small number of foreign mtDNA among total amount of mtDNA populations of the same genus. Therefore, interspecies or intergeneric mtDNA will be a useful tool to investigate the fate of foreign mtDNA in ooplasm.

Fate of interspecies or intergeneric mtDNA in ooplasm

SCNT offers the possibility of preserving endangered species. However, owing to the limited availability of oocytes from wild animals, the cloning of endangered species benefits by the use of recipient oocytes from a related domestic species. However, this methodology results in the production of nuclear–cytoplasmic hybrids [43, 44]. Despite numerous attempts in a wide variety of species, the number of live births of cloned offspring is still limited to instances that combine genetic compartments of closely related species, as observed in gaur–bovine [45], mouflon–sheep [46], African wild cat–domestic cat [47] and gray wolf–domestic dog experiments [48]. In most cases, interspecies/intergeneric SCNT (iSCNT) embryos developed to the blastocyst stage. Although 25.9 % of hybrid embryos in bovine (Bos taurus, 2n = 60) oocytes exposed to water buffalo (Bubalus bubalis) sperm reached the blastocyst stage [49], attempts failed to produce hybrid offspring [50, 51] with the exception of one rare case [52]. Similarly, buffalo–bovine iSCNT embryos demonstrated development to the blastocyst stage [53, 54, 55]; however, neither pregnancy nor natural delivery was reported. The generation of animals through iSCNT poses several problems, including mitochondrial/genomic DNA incompatibility, embryonic genome activation of the donor nucleus by the recipient oocyte, and availability of suitable foster mothers for iSCNT embryos. This orchestrated interaction requires the involvement of many nuclear‐encoded proteins with the displacement loop (D‐loop) of the mtDNA genome for efficient mediation of replication and transcription, regulation of total mtDNA copy number, and supply of ATP for energy‐requiring activities [56, 57, 58]. The species‐specific nature of mitochondrial biogenesis and function makes this modeling particularly significant for iSCNT. Quantitative analysis of mtDNA would be a powerful tool to study the interaction between donor nuclei and mitochondria after nuclear transfer. The success of gaur–bovine iSCNT illustrated potential “communication” between the gaur nuclear and bovine mitochondria by virtue of an increase in total mtDNA, including mostly bovine mtDNAs at the blastocyst stage [59]. In iSCNT embryos, the copy numbers of D‐mtDNA were found to be constant from the 1‐cell to the 8‐cell stage in water buffalo–bovine (Fig. 4) [60], sheep–bovine [61], cat–bovine [62], goat–sheep [63], and macaque–rabbit [64] arrested embryos. Similarly, in pronuclear bovine zygotes injected with (buffalo or murine) mitochondria illustrated that exogenous interspecies/intergeneric mtDNA injected into bovine oocytes were not selectively destroyed; indeed, they were maintained throughout early development (until the blastocyst stage) (Fig. 5) [65]. The drastic reduction of mtDNA content is in favor of active destruction rather than a reduced turnover of mtDNA molecules [23]. The need to exclude defective exogenous mtDNA from the developing embryo represents a strong selective pressure against survival of exogenous mtDNA. It is doubtful that exogenous intergeneric or somatic cell mitochondria would not be recognized by recipient cellular surveillance for mitophagy via the ubiquitin–proteasome pathway observed in sperm mitochondria. Transfer of buffalo ooplasm into bovine zygotes by fusion methodology introduced 8.3 % buffalo mtDNA and maintained a comparable ratio through to the blastocyst stage. However, no vestiges of buffalo mtDNA were found in subsequent offspring [66]. This would support an inability of buffalo mtDNA to establish a functional interaction with the bovine nucleus.

Figure 4.

MtDNA copy number per bovine SCNT and buffalo–bovine iSCNT embryo during early development [57]. Donor cell (buffalo, gray dotted line) and recipient oocyte (bovine, gray line) mtDNA in iSCNT embryos was stable during early development until arrested at the 8‐cell stage. Total mtDNA (bovine, black line) in SCNT increased at the blastocyst stage

Figure 5.

Bovine and buffalo mtDNA copy number per bovine oocyte/embryo injected with buffalo mitochondria. Oocytes/embryos injected with buffalo mitochondria (after injection) were analyzed after electric stimulation (after pulse), chemical activation by 6DMAP (after 6DMAP) and 7 days of culture [arrested at the morula, blastocyst, and Ex (expanded) blastocyst stage]. The black line represents bovine mtDNA copy number and the dotted line represents buffalo mtDNA copy number

Mitochondrial proteomic approach in cloned cattle and pigs

Mutations in the mtDNA result in developmentally regulated effects in animal models [67, 68, 69] and in characterized mitochondrial diseases in humans [57]. Previous reports provided evidence for the influence of oocyte mtDNA haplotype on SCNT efficiency in cattle [70, 71, 72]. The mitochondria transmitted by nuclear transfer may affect the phenotypic changes observed in cloned animals [73]. It is not easy to elucidate the underlying mechanism associated with this question. Mitochondrial heteroplasmy/exchange can affect developmental ability, health, and uniformity of SCNT offspring produced. Comparative proteomic analysis may be useful to identify differentially expressed mitochondrial proteins associated with animal cloning abnormalities and losses. Two‐dimensional differential gel electrophoresis (2D DIGE) is becoming more popular in protein quantification with the inclusion of an internal standard, providing a technology that greatly reduces experimental variation [74, 75]. Recently, differentially expressed proteins of SCNT piglets and placentae from SCNT pigs and cattle were reported using comparative proteomic analyses [76, 77, 78, 79]. We investigated the differences of protein expression profiles of mitochondria in adult SCNT pigs derived from Chinese pig (Meishan breed) fibroblasts and European pig enucleated oocytes (maternal Landrace breed) [17, 29] using a 2D DIGE system [80]. Although alteration of liver mitochondrial protein expression levels was observed in adult SCNT pigs, it was not reflective of breed differences associated with recipient oocytes. Interestingly, one of the SCNT pigs, derived from an SCNT embryo using an in vitro matured oocyte [81], showed the greatest profiling disparity compared with the other clones. In a similar study in cattle, mitochondrial protein profiles varied among adult clones produced from the same cell line as used for the donor cells [82]. One resultant SCNT cow, produced using an oocyte obtained from an ovary cultured overnight in PBS at 15 °C, showed the greatest profiling disparity compared with the other adult clones (Fig. 6). One possible explanation for this variation is that epigenetic differences led to the discordant outcomes in generating viable cloned animals.

Figure 6.

Protein expression changes of liver mitochondria of SCNT cattle analyzed by 2D DIGE (P < 0.05). Up‐regulated (>2.0) and down‐regulated spots (<−2.0) in each individual SCNT cattle (CA1–4; Fig. 1) were calculated versus the mean of the control group by the ImageMaster™ Platinum 6.0 software. ID: spot ID for calculation by software

Pre‐ and post‐natal development is often compromised, and a variable proportion of the SCNT offspring show aberrant developmental patterns and increased pre‐ and perinatal mortality [1]. Further, about half of surviving bovine cloned neonates frequently exhibit phenotypic abnormalities, including large‐offspring syndrome with both respiratory and metabolic deficiencies [83, 84]. Aberrant reprogramming of donor somatic cell nuclei may result in a variety of severe problems in animal cloning [2, 3]. Alterations in methylation levels, either globally or at specific sequences, were observed in abnormal and non‐viable bovine SCNT fetuses and calves, compared with either normal controls or putatively normal live‐born clones [85, 86]. Epigenetic differences may also influence expression of proteins in SCNT animals. Recently, comparative proteomic analyses were performed to qualify the key factors of SCNT animals’ abnormalities [76, 77, 78, 79, 87]. 2D gel electrophoresis analysis revealed changes in the responses of several detoxification‐related proteins to stress and inflammation observed in the sudden death of SCNT piglets with extensive hepatopneumonic apoptosis [87]. A host of differentially expressed proteins were identified in SCNT‐derived placentae [76, 77]. In term placentas from porcine SCNT, the expression of most of the detoxification‐related proteins and antioxidant enzymes was down‐regulated, and the global protein expression profiles illustrated that proteins closely involved in the apoptotic signaling pathway were different in comparison with controls [77]. Not only placental insufficiency, but down‐regulation of proteins involved in the prevention of oxidative stress, was observed, while molecules involved in apoptosis were up‐regulated in early post‐natal, but non‐viable, SCNT piglets [79]. Apoptosis related proteins with different expression patterns were also identified in liver mitochondria of post‐mortem SCNT calves [81]. Moreover, differentially regulated proteins were also observed in adult SCNT animals as compared to controls. Up‐regulation of proteins controlling glucose metabolism and ROS suggests that the accumulation of ROS may induce apoptosis in the pancreas in adult SCNT pigs [78]. Down‐regulation of the aldehyde dehydrogenase family 4 member A1, which acts as a negative regulator of p53‐dependent apoptosis [88], was found in liver mitochondria derived from adult SCNT pigs [80]. Similarly, an up‐regulated 78 kDa glucose‐regulated protein precursor, which enhances ER (endoplasmic reticulum) stress associated with cell death [89], was observed in liver mitochondria of adult SCNT cattle [82].

Although nuclear transfer is recognized as an effective technique for producing economically important and value‐added domestic animals, it has not reached an economically practical state. Cloned offspring were successfully produced in a variety of species, but SCNT technology still struggles with extremely low experimental efficiencies. Here, mitochondrial influences on animal cloning efficiency and success were detailed. However, further study is needed to clarify the intracellular characteristics of SCNT animals and to elucidate the cellular mechanisms that limit the practical use of this technique including basic biological question such as those related to epigenetics and nuclear–cytoplasmic interactions during development.

Acknowledgments

I thank Prof. C.A. Pinkert (Auburn University) for critical reading of this manuscript and Dr. H. Hanada, Dr. A. Onishi (NIAR), Dr. T. Kojima, M. Tasai, Dr. S. Akagi (NARO), K. Srirattana (Suranaree University of Technology), Dr. K. Nirasawa (NARO), and Prof. H. Imai (Kyoto University) for their help. Mitochondrial and SCNT studies were supported by a Grant from the NARO Japan.