Probability , odds and random chance : the difficult task of modulating the epigenetic profile of cloned embryos

Since the beginning of modern embryology, scientists have wondered about how a small number of totipotent embryonic cells can become an individual with a wide variety of organs and tissues with distinct functions. Also, the idea of generating a cloned animal using a nucleus from a donor cell is not recent. However, it has taken years of research to achieve this goal, especially regarding mechanisms of cell reprogramming required to return a differentiated cell to totipotency. Cloning by somatic cell nuclear transfer (SCNT) has been a valuable tool to understand epigenetic mechanisms related to cellular reprogramming. However, cloning efficiency is still low, with a low percentage of embryos resulting in healthy animals. The high attrition rate is associated with incomplete or abnormal epigenetic reprogramming, such that many cloned embryos have DNA methylation patterns different than controls, resulting in faulty gene expression and subsequent developmental failures. Attempts to improve genome reprogramming by modulation of oocyte quality and/or somatic cell plasticity, thereby increasing cloning efficiency and preventing detrimental effects on development, have proven ineffective. The recent development of DNA editing techniques may facilitate an improved understanding of cellular reprogramming and the role of DNA methylation in development. These novel tools may lead to new means to modulate epigenetic programming and inheritance, and hold great promise to assist in epigenetic remodeling of the donor nucleus. Such strategies are likely to improve the odds for successful cloning.


Introduction
Among many questions of biological significance, one in special has intrigued embryologists for a long time: how can cells change their phenotypes without changing the genotype?The concept of genomic equivalence, in which each somatic cell in the organism has the same genetic material and, therefore, also has all the information necessary to create a complete new organism, was assumed as a fact for decades, since cells are all derived from one single cell fertilized egg.However, the ultimate evidence came only in 1996 with the birth of Dolly, the sheep, cloned by somatic cell nuclear transfer (SCNT) procedures using somatic cells from an adult ewe (Wilmut et al., 1997).At last, animal cloning demonstrated that a differentiated somatic cell nucleus could be re-directed (or reprogrammed) for the development of an entire new organism.Since then, more than 20 vertebrate species have been cloned, from a variety of somatic cells, for a multitude of purposes, at various levels of success, demonstrating the usefulness of the procedure for applications ranging from science to the industry, animal conservation to production, genetic selection to transgenesis.
Twenty years after Dolly has passed, and cloning efficiency still remains low.Despite efforts to improve it, less than 5% of the animals come to term and remain healthy afterwards (Selokar et al., 2014(Selokar et al., , 2015)).Even though the precise factors are yet to be determined, such low efficiency is undoubtedly associated with incomplete or abnormal chromatin remodeling and genomic reprogramming after cloning, which is related to nonphysiological epigenetic modifications dependent on (i) the quality of the recipient oocyte or cytoplast where the donor nucleus need to be reprogrammed (Wells et al., 2010;Aguiar et al., 2016), (ii) the origin, quality, and plasticity of the donor nucleus or karyoplast per se prior and after its transfer into a cytoplast (Campbell et al., 2005), and (iii) the level of technical expertise and biological variation in all steps related to cloning (Bertolini et al., 2012).For normal embryo development and survival, success in cloning depends on correctly reversing specific markers of cellular differentiation to reactivate or resume processes needed during development, all associated with proper specific and global gene expression (Niemann, 2016).
Even with the concept of genomic equivalence, the ideia that the genome can be strongly affected by external or environmental influences lay the ground basis for what is referred as epigenetics, and what such a field is about: how genes are regulated, what factors can influence them, and how genes control and are controlled in the course of development throughout life, under physiological or pathological conditions, and in health and disease.In that regard, animal cloning still is a useful model to dedifferentiate specialized somatic cell nuclei to a totipotent state, in a process referred as epigenetic reprogramming.Thus, this review aims to briefly discuss some concepts and views on the role of epigenetic and reprogramming mechanisms in growth and development, especially after cloning by SCNT, including attempts to epigenetically modulate gene function and normal development.

Chromatin remodeling, epigenetic reprogramming, and early events in development
Nuclear remodeling after cloning or remodeling of chromosomal architecture is the designation for a series of molecular events involving modification and expression of nuclear laminins (Kubiak et al., 1993), formation of the nuclear envelope (Szollosi and Szollosi, 1988), and changes in protein synthesis (Fulka et al., 1996).The cytoplasmatic kinase named Mitosis or Meiosis Promoter Factor (MPF) has great importance in the nuclear remodeling (Campbell, 2002(Campbell, , 2013)).The kinase activation triggers the onset of mitosis or meiosis, resulting in nuclear envelope break (NEBD), chromatin condensation, reorganization of the cytoskeleton and changes in cell morphology (Nurse, 1990;Campbell, 2002Campbell, , 2013)).
It has not been determined if the nuclear remodeling is an absolute necessity for the normal development of cloned embryos.However, it has been demonstrated for quite some time now that nuclear MPF facilitates remodeling after cloning (Fulka et al., 1996).In metaphase II (MII) oocytes containing high levels of MPF active kinase, the maintenance of the cytostatic factor (CSF) that keeps MPF active remains until oocyte fertilization or activation (Fulka et al., 1996;Marteill et al., 2009).After the steps of fusion/activation in conventional cloning procedures, the kinase activity of MPF phosphorylates numerous cellular proteins leading to the completion of meiosis II, the remodeling of the maternal and paternal chromatin, also contributing to the embryonic activation process (Sirard, 2011).
In cloning, among the most notorious events of the MPF action in the MII oocyte on the donor nucleus, usually at the G0/G1 stages of the cell cycle, are the breakdown of the nuclear envelope, the premature chromatin condensation (PCC) and the formation of the achromatic spindle, which culminates in the extrusion of the second polar body (Campbell, 2013).Conversely, after fertilization, embryonic activation follows, when CSF is degraded and the MPF activity decreases in the first hours, coinciding with the nuclear envelope reconstruction in the form of male and female pronuclei and chromatin decondensation, initiating an interphase (Szollosi et al., 1988;Nakai, 2016).In cloned embryos, which are artificially activated by electrical or chemical processes, MPF activity remains high for a longer period of time (Kubiak et al., 1993;Sterthaus, 2009).Currently, researchers are still trying to identify factors and their importance in the success or failure of epigenetic reprogramming after cloning, seeking the understanding of how to improve protocols and processes to attain more reliable results.

Intrinsic and extrinsic modulators of epigenetic reprogramming and early development after cloning
Animal cloning by SCNT, although simple in concept, involves multiple steps that make the process complex, a fact responsible for its success or failure in subsequent development (Bertolini et al., 2012).Some steps, e.g., oocyte in vitro maturation (IVM) or in vitro culture (IVC) of embryos, are common to other in vitro embryo production (IVP) systems, such as in vitro fertilization (IVF) procedures.However, other steps are exclusive to cloning, such as those that replace fertilization by SCNT.
The cloning success is directly associated with the source, quality and preparation of enucleated recipient cytoplasms (cytoplasts), the donor cell nucleus (karyoplasts), and also cloning procedures per se, including embryo reconstruction and activation (Campbell et al., 2005;Wells, 2010).Therefore, the intrinsic features of cytoplasts and karyoplasts are modulated by physical and chemical factors, including manipulation and ex corpore exposure, which is always a suboptimal environment and biologically less suitable for development (Bertolini et al., 2012).

Cytoplast source and cell cycle synchrony/asynchrony
The oocyte source is a limiting factor for many species, in terms of availability, quality and biological competence, also being one of the main sources of variation in programs of in vitro embryo production (Galli et al., 2003;Wells, 2010).In vivo-matured oocytes are routinely used for cloning in many animal species, as such oocytes are presumably more competent for nuclear reprogramming and to subsequent embryonic development (Campbell et al., 2005;Ribeiro et al, 2009;Wells, 2010;Mezzalira et al., 2011;Aguiar et al., 2016).
The cytoplast appears to be one of the key elements for cloning success as everything associated with the oocyte is essential for embryonic and subsequent fetal development, still being one of the most efficient milieu for the induction of epigenetic reprogramming of the genome (Campbell et al., 2005).Reprogramming factors, for example, must be abundant and sufficient present to effectively reprogram a genome (Campbell et al., 2001;Oback and Wells, 2003;Wen et al., 2014).It is believed that nuclear reprogramming capacity reaches its highpoint from fertilization/reconstruction up to the embryonic genome activation, as the oocyte cytoplasm contains chromatin remodeling factors and reprogramming factors (proteins, mRNAs, miRNAs and molecular precursors) accumulated during folliculogenesis (Alberio et al., 2001;Urrego et al., 2015).However, factors acting in the reprogramming of the donor nucleus are still not completely identified, with the oocyte quality remaining key to successful animal cloning (Wells, 2010).
Pieces of evidence suggest that the start in embryonic genome expression occurs early in development (Flach et al., 1982;Graf et al., 2014), leading to the hypothesis that the nucleus requires a time to adapt to a new cellular environment.Once the oocyte environment is at a state of transcription, a transferred nucleus donor would not have adequate time to adapt, negatively affecting nuclear reprogramming.This leads to another important point of discussion, Anim.Reprod., v.14, n.1, p.102-123, Jan./Mar.2017 which is related to the level of synchrony between the cell cycles of the recipient cytoplasm and the donor cell.
It is known that the exposure of a donor nucleus to an environment with active MPF, such as in the MII oocyte, will lead to a series of subcellular events promoting exposure of chromatin reprogramming factors also at high levels in the cytoplasm, which possibly favors epigenetic reprogramming (Mezzalira et al., 2011).Thus, in vitro embryo development is improved when cytoplasts at MII are used in association with post-fusion activation (Heyman et al., 2002;Mezzalira et al., 2011).The time between the fusion of the cytoplast at MII with the karyoplast at G0/G1 and embryo activation is referred as reprogramming time, and such time appears to play an important role in chromatin exposure to oocyte cytoplasmic remodeling factors (Campbell et al., 2005).The appropriate reprogramming time usually ranges from 1 to 4 hours, but no consensus exists about the ideal exposure time under each separate condition.

MII oocytes as cytoplasts
Generally, classic cloning procedures, from Briggs and King (1952) to this day, use enucleated oocytes at MII as recipient cytoplasts.For such cloning protocols, the karyoplast cell cycle can be used either at the G1 phase, which is the interval between mitosis (M phase) and onset of DNA replication (S phase), or at the G0 phase (Wilmut et al., 1997;Kato et al., 1998;Heyman et al., 2002), with cell remaining metabolically active but mitotically quiescent (Oback and Wells, 2002).Then, the somatic cell nucleus must always adopt the parameters of the cell cycle of the MII recipient cytoplast (Kikyo and Wolffe, 2000).As non-activated MII cytoplasts have high MPF levels, donor nuclei are forced into the M phase, with cells at G0/G1 remaining diploid, whereas cells at the G2 or at the S phases result in aberrant chromosome configurations (Oback and Wells, 2002).In such cases, if the cloning protocol does not allow the extrusion of the second polar body, the resulting embryo may become tetraploid or, at its best, the rupture of the nuclear envelope and premature chromatin condensation (if DNA synthesis is not complete) can lead to aneuploidy and chromosome loss (Barnes et al., 1993(Barnes et al., , Østrup, 2009)).Then, cell cycle synchronization into G0/G1 is usually attained by many procedures to ascertain proper embryo development after cloning (Gerger et al., 2010).

Universal cytoplast
Pre-activated oocytes are referred as universal cytoplasts when used for cloning after the decrease in MPF activity, as such cytoplast is compatible with development using cells at any stage of the cell cycle.Under all situations, resolution of ploidy will occur, as there is neither the breakdown of the cell nuclear envelope at the interphase nor a premature chromatin condensation (Campbell et al., 1996).Nevertheless, the use of such oocytes usually results in lower rates of development.Similarly, aged oocytes have lowering CSF levels due to gradual decay, destabilizing MPF, leading to an increase in spontaneous activation by physicochemical environmental events (Wakayama et al., 2003).However, despite being easier to activate, aged oocytes have lower potential for further embryo development (Chian et al., 1992;Takahashi, 2013).

Late TII oocytes as cytoplasts
As an additional oocyte source for cloning, also related to the cell cycle, the embryo reconstruction using oocytes at late telophase II (Bordignon and Smith, 1998) results in proper embryo development, but at slightly lower rates than MII oocytes, possibly due to MPF levels still being abundant to trigger events observed with MII oocytes, some associated with reprogramming and chromatin remodeling (Mezzalira et al., 2011).Such system is successfully in use in commercial cloning operations in Brazil and abroad.

Karyoplast or nuclear donor cell type
The choice of the karyoplast or nuclear donor cell type has a significant biological impact on cloning and accounts for large variations in the success rates (Campbell et al., 2005, Liu et al., 2015).Cellular reprogramming efficiency after SCNT seems to depend on a multitude of factors, including the genotype (species, breed, gender, age, individual factor, etc.), the stage of development (embryonic, fetal, adult animal), tissue origin, cell type, degree of differentiation, cell cycle, and overall features when in culture (Batchelder et al., 2005;Wells, 2010).As mentioned above, the current understanding is that most of the cloning efficiency is associated with the use of synchronized somatic cells at G0 or G1 of the cell cycle (Wilmut et al., 1997;Wakayama et al, 1998;Wells, 2010).In addition, cells of embryonic origin seem to present a higher potential for further development and lower level of abnormalities than cells from fetal origin, which in turn seem to be more efficient than adult cells (Heyman et al., 2002;Batchelder et al., 2005).Cells from the same individual but from different tissue origins at distinct differentiation levels have also been described to favor cells of lower degree of differentiation (Batchelder et al., 2005).Likewise, the isolation method, the type of cell culture and the number of cell passages have an important effect on the results (Wells, 2010;Martins et al., 2016), whereas cells with few passages in culture appear to exhibit greater potential for epigenetic reprogramming after cloning (Li et al., 2014).All these aspects should be carefully controlled, as it is known that the culture conditions can cause genomic instability and consequent failures after cloning (Humpherys et al., 2001).Nonetheless, the small number of reported births, the absence of extensive controlled studies, and the low technical efficiency make it a difficult task to adequately compare cell types in cloning by SCNT (Tecirlioglu et al., 2005).

Embryo reconstruction procedures
The embryo reconstruction itself can be accomplished by micromanipulation with (Wilmut et al., 1997) or without (Oback and Wells, 2003) zona pellucida, or manually by handmade cloning (HMC) without zona pellucida (Vajta et al., 2003), with no apparent differences in the final efficiency between procedures (Tecirlioglu et al., 2005).Usually, embryo reconstruction by SCNT involves the steps of oocyte enucleation, nucleus transfer by membrane fusion with the mix of different cytoplasms, oocyte activation and embryo culture.Oocyte enucleation may result in loss of cytoplasm and, when excessive, can compromise subsequent embryo development (Ribeiro et al., 2009).On the other hand, the fusion of cell membranes leads to mosaicism and cytoplasmic heteroplasmy by mixing different cytoplasts, and it is still not clear which are the possible biological consequences on subsequent development (Ribeiro et al., 2009;Mezzalira et al., 2011).In addition, fusion aims to introduce a donor nucleus, typically at the G0/G1 phases of the cell cycle, into a MII oocyte.Such process exposes the chromatin to reprogramming factors and to heterologous proteins and miRNAs, all likely important for genome reprogramming (Alberio et al., 2001).
Oocyte or embryo activation also appears to be related to the possible development problems since, physiologically, it occurs by the action of factors released by the sperm cell, causing intracellular periodic oscillations in intracellular Ca 2+ ([Ca 2+ ] i ) in the oocyte for a certain period of time, which depolarizes the cell membrane and induces a chain of events that leads to embryo activation and the first embryo cleavage (Galli et al., 2003).In cloning by SCNT, activation of the reconstructed embryo is usually induced artificially by electric or chemical procedures, leading to elevated [Ca 2+ ] i in a single peak.Moreover, the use of inhibitors of protein kinases (targeting MPF) or protein synthesis (blocking cyclin translation), especially for cells at G0/G1, is deemed necessary for further embryo development (Campbell et al., 2005).Finally, the in vitro embryo culture system has also been implicated with epigenetic changes, resulting in distinct gene expression patterns in embryos and in embryonic stem cells (Dean et al, 1998;Wrenzycki et al., 2001).
Altogether, the summation of intrinsic and extrinsic factors related to the cytoplast source and karyoplast type, along with the biological and technical aspects in all steps involved in cloning procedures per se, ultimately will determine the overall cloning efficiency, which is by far and large dependent on the level of genome reprogramming at early embryo stages, making animal cloning a process of survival of the epigenetically fittest (Bertolini et al., 2012).

Improving genome reprogramming: is it feasible?
Overall, and despite such variety in the idiosyncrasies and peculiarities to reach success after cloning by SCNT, once a minimal understanding of the methodology, certain levels of skill and experise, and the control of some basic biological factors are attained, obtaining cloned blastocysts is a relatively easy outcome.Still, blastocyst yield does not necessarily reflect success.An example is the possibility of achieving high blastocyst rates by parthenogenesis, a procedure commonly used to infer oocyte quality (Ribeiro et al., 2009).Thus, although blastocyst yield is a rather easy pitfall to overcome, proof of success after cloning can only be certified when there is subsequent in vivo development after the embryo transfer (ET) to recipient females, culminating with the birth of viable offspring (Bertolini et al., 2012).
Pre-and postnatal abnormalities observed after cloning by SCNT are directly related to the efficiency of each step in embryo reconstruction.In the process, the functional and molecular asynchrony between donor nucleus and ooplasm components leads to inadequate genomic reprogramming.This fact is well demonstrated by several groups in which mammalian cloned embryos have an abnormal genome epigenetic configuration associated with phenotypic and physiological changes during development (Dean et al., 2001;Wrenzycki et al., 2001;Beaujean et al, 2004).Such features support the hypothesis that failures at the level of gene expression and embryonic development are caused by improper reprogramming of the donor genome during and after SCNT, due to many of the factors and components already discussed above, affecting the pattern of DNA methylation, or histone acetylation, methylation, phosphorylation or ubiquitinization during the first rounds of embryo development, compromising proper gene function and development afterwards.

Role of DNA methylation in development
DNA methylation in mammals is an important regulator of gene transcription, being a mechanism often used to silence and regulate genes without changing the original DNA sequence; this is one of the most important strategies for epigenetic modifications of the genome (Jaenisch and Bird, 2003).Biochemically, DNA methylation is a process involving the addition of a methyl group at position 5 of cytosine residues (5mC) in the DNA.In adult somatic cells, the DNA methylation typically occurs in dinucleotide sequences called CpG islands, which are in guanine-and cytosinerich regions.Between 60 and 90% of all CpGs are methylated in mammals (Tucker, 2001), and these islands are usually also present in the 5' regulatory regions of many genes, which are also associated with differentially methylated regions (DMR), directly related to genomic imprinting.
DNA methylation is essential for normal development and performs a variety of biological functions, being associated with very important processes, including repression of gene expression, genomic imprinting, X-chromosome inactivation, suppression of repetitive elements and carcinogenesis.Thus, the relationship between DNA methylation and chromatin structure is very important because this epigenetic modification affects cell function, allowing cells with the same genetic material to differentiate, yielding multiple organs, or perform various functions (Jaenisch and Bird, 2003).DNA methylation can also suppress the expression of endogenous retroviral genes and other deleterious DNA sequences that may have Anim.Reprod., v.14, n.1, p.102-123, Jan./Mar.2017 been incorporated into the host genome over time.The methylation can also protect DNA from enzymatic cleavage because most restriction enzymes are incapable of recognizing and binding to sequences (epigenetically) modified externally.
As one of the most important functions, DNA methylations at CpGs have the specific effect of reducing gene expression, being found in all vertebrates, with extensive evidence that genes with high levels 5methylcytosine in their promoter region are transcriptionally silenced (Jaenisch and Bird, 2003), as shown in Fig. 1.DNA methylation can affect gene transcription in at least two ways: 1) physically preventing the binding of transcription proteins; and/or 2) binding proteins known as MBDs (methyl-CpG-binding domain proteins).These proteins may recruit additional proteins to the site, such as histone deacetylases and other chromatin remodeling proteins that can modify histones, thereby forming a compact and inactive chromatin called heterochromatin (Nan et al., 1993).Interestingly, no significant methylation pattern is seen in embryonic stem cells (Lister et al., 2009), but in somatic cells there are high levels of DNA methylation generally transmitted to daughter cells with high fidelity.Therefore, the resulting epigenetic changes are normally stable, continuous and unidirectional, preventing an organism from reverting any level of cell differentiation or the conversion of a tissue to another.

Role of DNA Methyltransferases in development
The pattern of global and specific DNA methylation in mammalian cells appears to be propagated during the development of the conceptus after each DNA replication cycle by an action of either de novo or maintenance DNA methyltransferases (Jackson-Grusby et al., 2001;Golding and Westhusin, 2003).DNA methyltransferases (DNMTs) are part of a family of enzymes that catalyze the transfer of a methyl group to the DNA using S-adenosyl-methionine (SAM) as a methyl donor (Smith et al., 1992).Like restriction enzymes, methyltransferases have specific target DNA sequences.In mammalian cells, DNMTs methylate DNA sequences in CpG islands, located generally closer to or within the promoter regions, and exhibit different patterns of DNA methylation, also determining the imprinting pattern.In such cells, DNA methylation that occurs mainly at the C5 position in CpG dinucleotides is done by two general classes of DNA methyltransferases: maintenance and de novo DNMTs (Smith et al., 1992).Three active DNMTs have been identified in mammals, being called (maintenance) DNMT1, and (de novo) DNMT3a, DNMT3b and DNMT3L (Golding and Westhusin, 2003).

DNMT1
The DNMT1 is the most abundant DNA methyltransferase in mammalian cells and is considered the main maintenance enzyme, predominantly methylating CpG di-nucleotides in the hemimetilated genome (Kho et al., 1998).The maintenance of methylation activity is necessary to preserve DNA methylation responsible for the copy of the methylation patterns for "daughter" strands after each cellular DNA replication cycle.In vitro, the enzyme is 7 to 100 times more active in hemimetilated DNA compared to non-methylated substrate and is more active in the de novo methylation than other DNMTs.DNMT1 has several isoforms: the somatic form, the most common in the organisms and always present in the nucleus of somatic cells; variant DNMT1b isoform; and DNMT1o, an oocyte-specific isoform, which is synthesized and stored in the cytoplasm of oocytes, being allocated in the nucleus during specific stages of early embryo development (Golding and Westhusin, 2003).Abundant levels of transcripts to the main isoforms of DNA methyltransferases are found in oocytes and embryos at preimplantation stages (from 1-cell to blastocyst) in cattle (Golding and Westhusin, 2003) and humans (Huntriss et al., 2004).The DNMT1o variant, present in (a) (b) murine and human oocytes and embryos, is unidentified in cattle (Golding and Westhusin, 2003).

DNMT3
The DNMT3 is a DNA methyltransferase family that may also methylate the CpG sites of unmethylated and hemimetilated DNA.The structure is similar to DNMT1, with a regulatory region linked to a catalytic domain, with three known members of the family: DNMT3a, DNMT3b, and DNMT3L.The DNMT3a and DNMT3b can mediate methylationdependent gene repression, with these enzymes able to establish DNA methylation patterns in early embryonic development and to promote cell differentiation.The DNMT3L is a protein homologous to the other DNMT3s, but with no catalytic activity.Such protein interacts with DNMT3a and DNMT3b, being co-located in the nucleus, aiding the DNA binding to the de novo methyltransferases, stimulating their activity (Rhee et al., 2002).

Differentially methylated regions (DMRs) and genomic imprinting
The DMRs are genomic regions with different states of methylation, with functional regions involved in regulating gene transcription, and a strong association with genomic imprinting, including imprinting control regions (ICRs) into DMRs.The identification of DMRs from various tissues may show epigenetic differences among them, as well as cancer DMRs can show aberrant methylation patterns when compared with normal samples (Irizarry et al., 2009).It is well known that DNA methylation is associated with cell differentiation and proliferation (Reik et al., 2001), also strictly related to distinct DMRs in development and through the reprogramming process (Meissner et al., 2008;Doi et al., 2009).In addition, there are intra-and interindividual DMRs with longitudinal changes in the global methylation, increasing with age in a given individual or between individuals (Bjornsson et al., 2008;Bock et al., 2008).
The expression of imprinted genes crucial to development is also regulated at the epigenetic level by imprinting/methylation within DMRs, which regulate gene expression either from paternal or maternal alleles.One of the classical examples in imprinted gene regulation by imprinting/methylation within DMRs is played by the Igf2-H19 locus that encodes the pleitropic growth factor IGF2 and the H19 gene, which results in a non-coding RNA precursor of several microRNAs that have tumor suppressor effects (Reik and Murrel, 2000;Ratajczak, 2012;Singh et al., 2012), as illustrated in Fig. 2.Under normal imprinting pattern, a DMR positioned between the Igf2 and H19 coding regions is unmethylated on the maternal allele, and methylated on the paternal counterpart, with the expression of both genes being under the regulation of a downstream 3'distal enhancer (Fig. 2a).The methylation of the paternal DMR prevents binding of the CTCF insulator protein, allowing the activation of the paternal Igf2 promoter by the distal enhancer.Conversely, as the maternal DMR is unmethylated, CTCF insulator protein binds to it, preventing the activation of the maternal Igf2 promoter, with the maternal H19 being transcribed.Then, the proper somatic imprint of the DMR at the Igf2-H19 locus, with a lack of methylation on the maternal allele and a methylation pattern of the paternal allele, ensures that Igf2 is transcribed only from the paternal chromosome, and H19 only from the maternal chromosome, under a balanced expression pattern (Reik and Murrel, 2000;Ratajczak, 2012;Singh et al., 2012).
However, under certain epigenetic conditions, such as after faulty nuclear reprogramming following cloning by SCNT or in primordial germ cells (PGCs), the pattern of expression may change significantly, affecting pre-natal growth, development and survival to term.In the erasure of imprinting, when both DMRs at the Igf2-H19 locus are unmethylated, only H19 is transcribed from both parental alleles (Ratajczak, 2012).Conversely, in the loss of imprinting, when both DMRs at the Igf2-H19 locus are methylated, as seen in cancer cells in several cancer types, only Igf2 is expressed from both parental chromossomes (Fig. 2b and c).The imprinting process appears to be more susceptible to environmental effects and faulty reprogramming, with deregulations of ICRs/DMRs and loss of imprinting, generally resulting in abnormal phenotypes (Rancourt et al., 2013).Consequently, embryonic manipulations, such as IVF and cloning by SCNT, should deregulate genomic imprinting (Moore, 2001).

Physiological epigenetic changes in the course of development
DNA methylation appears to be critical to the course of normal embryonic development in mammals, as chromatin remodeling after fertilization is closely related to a fast demethylation of the parental genome (Reik et al., 2001;Dean et al., 2003).As illustrated in Fig. 3, DNA methylation pattern in gametes is typically removed during early embryo stages, to be restored during the successive cell divisions in development, altering the stable gene expression during cell division and inducing cell differentiation.In fact, global genomic reprogramming occurs rapidly after fertilization, with the degree of DNA methylation decreasing in about 30% of the average level observed in somatic cells (Bird, 2002).This fact is interesting because when demethylation was experimentally induced below such values, a negative interference in embryonic development was observed, showing that a certain degree of methylation needs to be maintained in the genome (Giraldo et al., 2009).Following the reduction in DNA methylation during the first cleavages, a de novo methylation pattern is promoted by DNMT3a, 3b and 3L (Fig. 3), beginning at the 8-cell and at the blastocyst stages in cattle and mice, respectively (Dean et al., 2003).In the mouse, methylation of the germinal vesicle nucleus increases with size of the oocyte.After fertilization, a quick hydroxymethylation of the paternal genome (blue), previously thought to occur as an active demethylation process, and a passive demethylation of the maternal genome (red) occurs.De novo methylation occurs at the blastocyst stage, with a differential methylation pattern between the inner cell mass (ICM) and the trophectoderm (TE).In cloned embryos (purple), some demethylation takes place by the blastocyst stage, after nuclear transfer but before de novo methylation, with aberrant methylation (hypermethylation) of the TE.From Yang et al. (2007).

Hydroxymethylation
Until recently, the demethylation of the paternal genome was considered as part of an active process (replication independent), whereas in the maternal genome, the demethylation occurs passively during DNA replication (replication dependent) in each cell division (Yang et al., 2007).Pieces of evidence suggest that in fact a passive process also occurs in the paternal genome, by the convertion of 5'methylcytosine (5mC) into 5'-hydroximethylcytosine (5hmC) prior to demethylation per se (Wossidlo et al., 2011).Enzymes previously linked to chromosomal translocations in leukemia (Lorsbach et al., 2003), the ten-eleven translocation proteins (TET1, 2 and 3), have been recently implicated as responsible for the oxidation of 5mC into 5hmC (Kriaucionis and Heintz, 2009;Tahiliani et al., 2009).Then, the final demethylation process is resolved by enzymes of the DNA repair machinery in the cell (Fig. 4).
The drastic drop in paternal DNA methylation after fertilization (Mayer et al., 2000), attributed earlier to a fast and active demethylation, occurs concurrently with an increase in conversion from 5mC to 5hmC (Wossidlo et al., 2011).Conversely, at the maternal genome, an aparent partial "protection" against TET3 prevents the addition of hydroxyl groups in the DNA, leading to a passive demethylation pattern (Nakamura et al., 2012).The mechanism used to control the methylation assumes that 5hmC interferes with the recognition of methylated cytosines by DNMT1, allowing a passive demethylation pattern to be established (Fig. 4a).Another possible path is explained by the low affinity of the 5hmC to methyl-CpG binding proteins, which specifically recognize the hydroxymethyl group, e.g., DNA repair proteins as 5hmC-specific DNA glycosylase (5hmC-DG), as shown in Fig. 4b (Dahl et al., 2011).

Abnormal patterns of DNA methylation and consequences for development
Chromatin remodeling after fertilization, associated with the reprogramming of the DNA methylation, appears to occur similarly in embryos produced both in vivo and in vitro (Han et al., 2003).However, several SCNT cloning studies showed that the introduction of the somatic donor nucleus (highly methylated) in a recipient cytoplasm (oocyte) filled with chromatin remodeling components is not sufficient to reprogram, modify or delete certain stable genomic epigenetic markers of differentiation, leading to faulty gene expression patterns (Bird, 2002).This reprogramming failure is associated with the hypermethylation status of DNA in murine and bovine embryos produced by SCNT, which may have similar levels of DNA methylation of the somatic cell donor nucleus (Dean et al., 2001(Dean et al., , 2003;;Beaujean et al., 2004).This fact can be observed in Fig. 3 above, where a high amount of methylated DNA is maintained after nuclear transfer in mice, while in vivo-derived embryos go through normal processes of replication dependent and independent demethylation.
Due to its role in gene expression, epigenetic modification is considered essential to the memory of a specific cellular function during development (Bird, 2002) and regulation of nuclear reprogramming (Han et al., 2003).The efficiency of epigenetic reprogramming after cloning by SCNT seems to be strongly dependent on the type and state of differentiation of the somatic cell donor used (Wells et al., 2003).In rats and cattle, the type of cell lineage in culture, a cell line or even subclone populations may reveal different developmental capabilities prior to and after cloning (Humpherys et al, 2001;Vichera et al., 2013).
An apparent relationship exists between the process of chromatin remodeling, genome reprogramming, and the profile of DNA methylation.In mice, a significantly higher percentage of blastocysts are obtained when cloned embryos are reconstructed with rather unmethylated cells, such as murine embryonic stem cells (ESC) or primordial germ cells (PGC), than other somatic cell types (Humpherys et al., 2001).In the absence of DNMT1 in the nucleus and after fertilization, the replication generates only unmethylated DNA strands which, over time, leads to passive DNA demethylation, as occurs with zygote DNA from maternal origin.Interestingly, higher amounts of transcripts for DNMT1 were observed in bovine blastocysts derived by cloning SCNT (Wrenzycki et al., 2001), which coincides with the detection of a more methylated state of cloned embryos.Thus, it is possible that the hypermethylated state of cloned bovine embryos is associated not only from poor demethylation of nuclear DNA, but also through the presence of active DNMT1 of somatic origin in the nucleus during early development, contributing to hypermethylation.
Genomic imprinting appears to be more susceptible to epigenetic changes, with deregulation of ICRs or DMRs generally resulting in abnormal phenotypes (Rancourt et al., 2013).Consequently, embryonic manipulations, such as IVF and SCNT, seem to deregulate imprinting (Moore, 2001).On the other hand, when DNMT1 activity is reduced, the DNA methylation in ICRs or DMRs associated with genomic imprinting was generally refractory to epigenetic changes, maintaining their methylation patterns in cells in culture (McGraw et al., 2013).However, the permanent loss of DNMT1 by gene deletion leads to a reduction in monoallelic expression of several imprinting genes (McGraw et al., 2013).Interestingly, homozygous DNMT1 mutant mouse embryos null by deletion died between days 9-11 of gestation (Jackson-Grusby et al., 2001).The loss of DNMT1, however, had no effect on the proliferative potential of mouse ESC in culture, but the induction of differentiation in these cells also lead to cell death (Rhee et al., 2002).On the other hand, DNMT1 overexpression induces a progressive DNA hypermethylation associated with transcriptional inactivation of several imprinting genes, leading to embryonic lethality (Biniszkiewicz et al., 2002).
It is not surprising that the set of distinct epigenetic features and changes at the transcript levels observed in cloned bovine blastocysts (Wrenzycki et al., 2001) should cause a significant effect in the expression of a variety of imprinting and nonimprinting genes after cloning (Wrenzycki et al., 2001;Bertolini et al., 2002Bertolini et al., , 2004)).Thus, an inadequate activation or inactivation of important developmentally important genes can predispose animals to different degrees of abnormalities after cloning by SCNT, depending on the interaction between qualitative and quantitative discrepancies, subsequent to the level or degree of genome reprogramming obtained after embryo reconstruction (Bertolini et al., 2012).For instance, cells with changes in imprinted genes may be directed to the trophectoderm of cloned embryos, potentially generating defective fetal membranes, as frequently observed after cloning (Wei et al., 2011).
The modified genome configuration of cloned embryos by aberrant DNA methylation patterns may be closely associated with reprogramming failures, altered gene expression profiles, and abnormal embryo development.In cattle, abnormal DNA methylation patterns have been linked to abnormalities after cloning, which in turn can be phenotypically expressed by fetal and placental disorders, increased pre-natal losses and lower postnatal survival (Bertolini and Anderson, 2002;Han et al., 2003;Salilew-Wondim et al., 2013).However, a certain degree of variation in gene expression pattern after cloning, generally attributed to epigenetic features transmitted by the donor cell nucleus, is totally acceptable and compatible with subsequent normal growth, development and survival after birth (Humpherys et al., 2001;Wells et al., 2003).

Manipulation of the epigenetic profile in the somatic
cell or the early embryo: there is no silver bullet

Epigenetic therapy
Currently, the manipulation of cellular epigenetic profiles, such as DNA methylation levels or histone modifications, has been applied in epigenetic therapy (Daura-Oller et al., 2009).Since the birth of Dolly, the scientific community and the commercial users have been coping with the low efficiency observed after cloning by SCNT, along with the appearance of abnormalities in the course of development, usually referred as the Abnormal Offspring Syndrome (AOS).As abnormal gene expression patterns appear to be a consequence of faulty epigenetic reprogramming after cloning, several potential strategies for improving cloning outcome have been endeavored for use in gametes, somatic cells and/or early developing embryos of several animal especies, as illustrated in Fig. 5 and Tables 1 and 2. Efforts have been dedicated to modulate key cellular events to mold or mimic physiological processes to obtain more suitable cytoplasts and karyoplasts for epigenetic reprogramming, chromatin remodeling, and related biological processes aiming to support proper development pre-and post-natal development after cloning by SCNT.Such attempts to modify epigenetic features of somatic cells or embryos, either at specific targets (DNA methylation, histone modifications, X chromosome inactivation, Table 1) or for broader alternative epigenetic modifications and/or on processes to enhance embryo activation and development (Table 2) have shown variable results.The basis for the most applied or promising strategies for the epigenetic modulation of mammalian cell nuclei is discussed below.

DNA demethylating strategies
In oncology, many tumor suppressor genes are silenced by DNA methylation and several chemotherapeutic treatments have been developed to reduce genome methylation in an attempt at reexpressing suppressor genes.Thus, to reduce the DNA methylation levels in cancer cells, demethylating agents, such as the cytidine analogues 5'-aza-cytidine (5-Aza-C) or 5'-aza-2'-deoxycytidine (5-Aza-dC), have been used in chemotherapy to inhibit DNA methylation by DNMTs, as the cytidine has a nitrogen at the position 5, instead of a carbon.However, to work, such agents must be incorporated into the genome, which can cause mutations in daughter cells.Some of such strategies using 5-Aza-C and 5-Aza-dC have also been attempted for cloning, with wide-ranging results.Generally, an adverse effect on embryo development is seen after cloning due to exposure of donor cells or embryos to relatively high doses of the cytidine analogues for rather long periods.Therefore, even if used with some success as chemotherapeutic agents in oncology, the demethylating chemicals above are unstable and may promote toxic and irreversible long-term effects.These treatments may adversely affect the period of chromatin remodeling of the embryonic genome, which includes the remethylation after the 8-cell stage in cattle (Dean et al., 2003).Other demethylating agents, such as adenosyl homocysteine (SAHA) and zebularine, more stable and less toxic DNMT inhibitors than the previous (Yoo et al., 2004), in addition to other demethylating strategies, including the RNAi-induced depletion of DNMT1 or the overexpression of cytidine deaminase, have been also applied to cells and embryos with promising results.

Increasing histone acetylation
When acetylation occurs in the histone structure, it becomes less compact, being more permissive to DNA transcription.The use of some chemical agents have been rather successful to increase histone acetylation, mainly by reduce the effects of deacetylase enzymes, aiming at improving cloning efficiency (Table 1).Trichostatin A (TSA) was demonstrated as a potent histone deacetylase inhibitor (HDACi), with greater specificity than other reagents previously used (Yoshida et al., 1990).Since then, TSA has been used to enhance embryo production efficiency by SCNT, as its use increases the amount of acetylated histones and was also related to a decreased histone methylation pattern.Scriptaid is a HDACi structurally similar to TSA, although it is considered more effective and less toxic (Su et al., 2000).Novel HDACi, such as oxamflatin, has been also used in pig embryos, but more studies are needed to verify its efficacy after cloning (Hou et al., 2014).

Histone demethylating strategies.
Histone 3 methylation of lysines at positions 9 (H3K9) and 27 (H3K27) is associated with transcriptional repression, with H3K9me3 being highly correlated with constitutive heterochromatin, which is the condensed, transcriptionally inactive state of chromatin.Recently, histone demethylating modulators, such as BIX-01294 or histone methylases, have been successfully used to increase the overall SCNT cloning efficiency.BIX-01294 inhibits the G9A histone lysine methyltransferase (HMTase), playing a role in H3K9 methylation, regulating gene expression and chromatin organization (Kubicek et al., 2007).Bovine somatic cells and one-cell embryos Use of 5-aza-dC when with TSA provided higher blastocyst rates and showed significantly lower levels of DNMT1,DNMT3b, HDAC2, and IGF2 transcripts Wang et al., 2011 Swine somatic cells and onecell embryos Increased blastocyst rates and OCT4, SOX2, NANOG expression when used in fetal fibroblast cells Huan et al., 2013 Buffalo somatic cells Reduced global methylation and abundance of DNMT1; HDAC1 decreased when with TSA, also, the blastocyst rates were higher using cells treated with both reagents  The use of cell-free systems allows the manipulation of cell programs, by the exchange of nuclear factors between distinct cell types.Fractionated cell homogenates (cell-free extracts) can maintain a biological function and are an important tool for the study of cellular and molecular processes (Alberts et al., 2002), providing a valuable means for examining biological mechanisms associated with nuclear and chromatin remodeling (Kanka, 2003).In fact, the ability of somatic cell extracts to modulate epigenetic reprogramming and chromatin remodeling, activating gene expression in differentiated somatic nuclei have been shown for years.Using this technology, Hakelien et al. (2002) and Landsverk et al. (2002) developed a somatic-cell-free system (stimulated human T cell extracts) that remodeled chromatin (hyperacetylation of a specific repressed gene promoter) and activated the expression of repressed genes in heterologous differentiated nuclei (293T fibroblasts and resting human T cell nuclei).Such studies paved the way for further studies on strategies using cell fusion, cell extracts or specific cell factors for the treatment of somatic cells or early embryos for broader-spectrum epigenetic modifications or to enhance embryo activation and development, as illustrated in Table 2. Overall, such systems also have variable results, lacking simplicity and effectivity.However, the identification of factors with biological functions usually become natural candidate molecules to target specific epigenetic modifications in somatic cells or early embryos for cloning by SCNT.
The present and the future: RNA-guided epigenetic regulators

RNA interference (RNAi)
The process of RNAi is an important posttranscriptional silencing mechanism, conserved in mammalian cells, oocytes and embryos, also significant for the maintenance of genome stability and early development (Svoboda et al., 2004).The RNAi technology has been effectively used for specific mRNA depletion from both mother and zygotic source (Svoboda et al., 2004), providing an efficient approach to induce a transient downregulation of developmentally important molecules, such as DNMTs, either in somatic cells or embryos, without inducing residual effects on the subsequent development (Table 1).Since this technology is potentially less toxic and without prolonged direct residual effect, RNAi can be used as an alternative to the chemical processes used to modulate embryonic reprogramming.

DNA editing and DNA regulation tools
Programmable nucleases, such as Zinc Finger Nucleases (ZFNs), Transcription Activator-like Effector Nucleases (TALENs) and Clustered Regularly Interspaced Short Palindromic Repeat (CRISPR)/ CRISPR-associated protein (Cas9), have been developed for effective gene editing, and have been broadly used in mammalian organisms (Gaj et al., 2013;Doudna and Charpentier, 2014;Kim and Kim, 2014).The CRISPR/Cas9 system is the most powerful technology and the main gene editor available, due to its versatility and design simplicity (Whitelaw et al., 2016).Yang (2015) has recently demonstrated the power of such tool by performing a one-shot genome-wide inactivation of 62 porcine endogenous retroviruses (PERVs) using a multiplex system.In such way, cells or embryos can be genetically modified either by sitedirected mutations or by gene insertions, generating transgenic founders potentially at a greater efficiency than using traditional methods.In fact, generation of specific double strand breaks (DSBs) by CRISPR/Cas9 has been exploited to modify the genome of livestock species, either through the non-homologous end joining (NHEJ) repair machinery, mostly as gene repressor to disrupt functional alleles (Hai et al., 2014;Ni et al., 2014;Crispo et al., 2015;Wang et al., 2015, Bevacqua et al., 2016), or through the homologous recombination (HR) pathway, for precise gene insertion (Ruan et al., 2015;Jeong et al., 2016;Wu et al., 2016).One example of the usefulness of the CRISPR/Cas9 system in epigenetic reprogramming out of a multitude of possibilities is the knockout of genes coding for histone methyltransferases associated with repressive states of the chromatin or heterochromatin, which will favor gene expression.However, in addition to the gene-editing role, the system CRISPR/Cas9 can also be used as epigenetic repressor or activator without editing the genome, which promises to become the Holy Grail in epigenetic reprogramming of cells, tissues or whole organisms in the near future.

CRISPR interference (CRISPRi)
Recently, the CRISPR/Cas9 system has gained a new version, the nuclease deactivated Cas9 (dCas9), that blocks transcription in prokaryotic and eukaryotic cells, known as CRISPR interference (CRISPRi), which can be used for the purpose of gene regulation instead of genome editing.As this catalytically dead Cas9 lacks endonuclease activity, it can be used to knock down multiple target genes simultaneously, with reversible effects (Qi et al., 2013).When coexpressed with a guide RNA, CRISPR/dCas9 generates a DNA recognition complex that can specifically interfere with transcriptional elongation, RNA polymerase binding, or transcription factor binding.This new system presents some advantages over the use of RNAi, mainly because the latter causes incomplete silencing of the desired gene (Jackson et al., 2003, Krueger et al., 2007), becoming an excellent alternative to transcriptionally repress genes, moving toward the study of different metabolic pathways in cells and embryos of different mammalian species.

CRISPR activator (CRISPRa)
In addition to the CRISPRi, the CRISPR/Cas9 system can be tailored to epigenetically modify the chromatin at specific target sites in the genome, activating silenced genes (CRISPR activating system, or CRISPRa).Such application can benefit the treatment of human epigenetic diseases, such as the Angelman Syndrome, in which a functional parental allele is silenced due to genomic imprinting, whereas the other allele is mutant.In such cases, it is possible to use the CRISPR/Cas9 system as an epigenetic activator of silenced genes, allowing expression of an intact copy of a missing gene (Vora et al., 2016).For instance, Hilton et al. (2015) fused Cas9 to the core catalytic domain of the histone acetyltransferase (HAT) of the human E1Aassociated protein p300, which allowed the acetylation of the lysine residue at position 27 of Histone 3 (H3K27), activating gene expression.Such tool will soon be also available for other epigenetic applications, as for instance, the use of fused Cas9-demethylases to perform precise and specific DNA demethylation in the genome.

Final considerations and perspectives
In summary, no silver bullet is available for the precise and complete epigenetic reprogramming of cell nuclei for any given purpose or application.Also, a solid relationship between in vitro treatments and in vivo efficiency is yet to be established (Sangalli et al., 2012;Kim et al., 2015).Moreover, no consensus exists on the usefulness of such epigenetic modulating systems mainly due to differences between species, cell types, doses, combinations, and treatment times between studies, among biological and technical factors, which makes such strategies of limited application.
It is apparent that we are just beginning to understand the epigenetic mechanisms in cell biology, which is the basis for improving genome reprogramming and cloning success.We are also only scratching the surface for the effective and widespread use of novel technologies, especially the RNA-guided epigenetic modulators.Irrespective of the cloning outcome, and due to the low efficiency of the process per se, progress in the field depends and will rely on the development of effective ways to reversibly modulate the epigenetic marks in the donor nucleus by inducing more permissive chromatin states suitable for reprogramming and/or to improve the reprogramming internal milieu in cytoplasts prior and after cloning.
As mankind aquires more knowledge into epigenetics, and as procedures to manipulate gametes, cells and embryos evolve to mimic nature into more homeostatic and homeorhetic physiologic processes, all fields in biology and medicine will advance towards the resolution of problems related to epigenetic programming and inheritance, minimizing random chance, thus improving the odds ratio for successful cloning.

Figure 1 .
Figure 1.Regulation of gene transcription by DNA methylation.(a) Active transcription: unmethylated CpG islands at promoters of genes leads to active transcription.(b) Inactive transcription: hypermethylated CpG islands at promoters of genes leads to transcriptional repression.TF: transcription factors.RNA Pol: RNA polymerase.From Turunen and Ylä-Herttuala (2011).

Figure 2 .
Figure 2. Regulation of expression pattern in the imprinted genes Igf-2 and H19 by changes in the methylation state in the DMR within the Igf-2-H19 locus, and expression balance (proportion) between genes.(a) Proper somatic imprint.(b) Erasure of imprint.(c) Loss of imprint.See text for details.M: maternal chromosome.P: paternal chromosome.CTCF: CTCF insulator protein.DMR: differentially methylated region.From Ratajczak (2012).

Figure 3 .
Figure3.Methylation dynamics during normal early embryonic development and in clones.In the mouse, methylation of the germinal vesicle nucleus increases with size of the oocyte.After fertilization, a quick hydroxymethylation of the paternal genome (blue), previously thought to occur as an active demethylation process, and a passive demethylation of the maternal genome (red) occurs.De novo methylation occurs at the blastocyst stage, with a differential methylation pattern between the inner cell mass (ICM) and the trophectoderm (TE).In cloned embryos (purple), some demethylation takes place by the blastocyst stage, after nuclear transfer but before de novo methylation, with aberrant methylation (hypermethylation) of the TE.FromYang et al. (2007).

Figure 4 .
Figure 4. (A) Top: DNA methyltransferase 1 (DNMT) recognizes 5mC and maintain methylation during DNA replication.Bottom: 5hmC is not recognized by DNMT, which will prevent maintenance methylation during DNA replication, resulting in passive (replication dependent) DNA demethylation.(B) Top: 5mC is not recognized by DNA repair proteins, maintaining methylation during DNA replication.Bottom: 5hmC may be recognized by DNA repair proteins, e.g., a 5hmC-specific DNA glycosylase (5hmC-DG), which will convert 5hmC to cytosine, leading to active (replication independent) DNA demethylation.Adapted from Dahl et al. (2011).

Figure 5 .
Figure 5. Potential strategies for improving the cloning outcome by the modulation of cellular events in gametes, somatic cells and/or early developing embryos for tailoring more suitable cytoplasts and karyoplasts for epigenetic reprogramming, chromatin remodeling, and related biological processes to support proper development pre-and post-natal development after cloning by SCNT.

Table 1 .
Reprogramming strategies to target specific epigenetic modifications in somatic cells or early embryos for cloning by SCNT.

Table 2 .
Examples of strategies for the treatment of mammalian somatic cells for cloning by SCNT or early embryos for broader epigenetic modifications or to enhance embryo activation and development.