| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Reproductive Technology |
a Infigen, Inc., DeForest, Wisconsin 53532
b Department of Cell and Molecular Biology, Howard Hughes Medical Institute, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104
c Department of Animal Sciences, University of Wisconsin-Madison, Madison, Wisconsin 53706-1284
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
embryo, developmental biology, early development, gene regulation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Recent successes in animal cloning using nonembryo-derived cells [5, 6] suggest that, for proper development to ensue, a donor nucleus must undergo a reversal in differentiation and a genome-wide epigenetic reprogramming. Based on observed abnormalities in some cloned animals, the lack of proper epigenetic reprogramming has been postulated to be an underlying factor contributing to improper development [7–9]. These anomalies include high abortion rates, placental abnormalities, increased birth weight, and perinatal death and have been reported in several species, including mouse [7, 8, 10], sheep [11], and cattle [12, 13]. Moreover, several abnormalities observed in cloned fetuses and offspring are strikingly similar to developmental irregularities attributed to aberrant expression of imprinted genes [14]. Imprinted genes are subject to epigenetic regulation by DNA methylation [15, 16], and it has been proposed that improper epigenetic reprogramming of imprinted genomic loci would account for aberrant phenotypes [8]. For example, elevated IGF2 (insulin-like growth factor-2) concentrations are associated with fetal overgrowth in mice [17] and changes in IGF2R (insulin-like growth factor-2 receptor) expression have been observed in large sheep offspring [18]. Moreover, altered expression of IGF2R in sheep is associated with decreased DNA methylation. Interestingly, however, mice cloned from embryonic stem (ES) cells demonstrate considerable variation in expression of imprinted genes [8] and yet postimplantation development of ES-cell NT embryos appears to be more efficient than somatic cell-generated nuclear transfer (NT) embryos [19].
Aberrant DNA methylation patterns have been reported in genomic repetitive elements in cloned bovine blastocysts [20–22] in which increased methylation of a CpG-rich satellite I repetitive element is observed when compared with in vivo and in vitro fertilization (IVF) controls [21]. In contrast, in cloned pig embryos, demethylation of centromeric satellite and the PRE-1 short interspersed element (SINE) occurs as it also does in IVF- and in vivo-derived embryos [23]. It remains unclear, however, how differentiated cells are epigenetically reprogrammed and how genomic methylation patterns are established during embryogenesis and development. Moreover, it is unknown to what extent the epigenetic modification of DNA methylation, which occurs in normal development, needs to be mimicked for nuclear transfer to succeed.
To gain a better understanding of global methylation changes associated with development of NT-generated cattle, we analyzed the genome-wide methylation status of cloned fetuses, aborted cloned fetuses, and adult clones by reverse-phase HPLC.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Tissue samples were incubated overnight at 55°C in 500 µl of extraction buffer (1 M Tris-HCl, pH 8.0, 5 M NaCl, 0.5 M EDTA), 12.5 µl 20% SDS (sodium dodecylsulfate) and 5 µl Proteinase K (Qiagen). The samples were extracted twice with phenol:chloroform (1:1) and precipitated with one volume of isopropanol. Following a wash with two volumes of 70% ethanol, the DNA was resuspended in 500 µl of TE (10 mM Tris-HCl, pH 7.5, 1 mM EDTA). All animal work was performed under Institutional Animal Care and Use Committee guidelines and standard animal-use protocols.
Reverse-Phase HPLC Analysis
Twenty-five micrograms of genomic DNA were diluted in dH2O to 50 µl, followed by denaturing at 95°C for 5 min and placement on ice. DNA samples were then incubated at 37°C for 2 h in 100 µl of 30 mM ammonium acetate (pH 5.3), 5 µl 20 mM ZnSO4, and 10 µl nuclease P1 (1 mg/ml in 30 mM ammonium acetate, pH 5.3). Following this incubation, samples were added to 20 µl Tris-HCl (pH 8.5) and 15 µl calf intestinal alkaline phosphatase and incubated at 37°C for 2 h. Digested samples were stored at -20°C prior to HPLC analysis. HPLC analysis was performed at the University of Wisconsin Biotechnology Center as described elsewhere [24]. All experiments were conducted in triplicate. Briefly, 50 µl of the digested sample were injected into a Brownlee Lab Spheri-5 RP-8 column (Alltech, Deerfield, IL) and separated at a flow rate of 0.75 ml/min. The gradient program involved 30 min in buffer A (0.05 M potassium phosphate, pH 4.0, 2.5% methanol) and 19 min in buffer B (0.05 M potassium phosphate, pH 4.0, 20% methanol). The column was then flushed with 70% ethanol for 13 min and reequilibrated with buffer A for 23 min prior to injection of the next sample. All samples were analyzed on a Beckman Instruments Gold Chromatograph, and nucleosides were detected at A280. Standard curves and peak positions to determine nucleoside concentrations were generated using nucleoside standards (Sigma). The percentages of 5-methylcytosine (5mC/(5mC + C)) in fetal and adult genomes were calculated from integration peak areas for each genomic nucleoside relative to peak areas of standard nucleosides (Sigma) of known concentration, detected at A280. Differences among cloned and in vivo fetuses and cloned and in vivo adults were assessed using a one-tailed Student t-test. Variances were analysed according to Levene's test, and were equal (P > 0.10).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
60–75-day-old fetuses and two 6-mo-old and two 7-mo-old fetuses), seven cloned fetuses recovered from ongoing normal pregnancies (60 days), four IVF fetuses (45–60 days), six fetuses generated by artificial insemination (AI; 60 days), 11 cloned adults, and four adults generated by AI. Genomic DNA was enzymatically digested into single nucleotides and processed according to a standard HPLC protocol [24]. The percentage of 5-methylcytosine relative to total cytosine 5mC/(5mC + C) in fetal and adult genomes was calculated by integrating the peak areas for each genomic nucleoside relative to the peak areas of standard nucleosides (Sigma) of a known concentration. The results indicated an extensive hypomethylation of DNA in the spontaneously aborted clones. The average 5mC content of these genomes was 7% (Table 1). Strikingly, six of the nine spontaneously aborted cloned fetuses had no detectable level of cytosine methylation despite differences in fetal development stages, ranging from 60 days to 7 mo (Fig. 1, A to C; Table 2). The remaining three clones, aged 180, 210, and 60 days, had 21%, 24%, and 24%, respectively. This finding was also supported by methylation-sensitive restriction endonuclease digests of genomic DNA with the isochizomers HpaII and MspI (data not shown), which suggested extensive hypomethylation in aborted cloned fetuses compared with cloned fetuses from ongoing pregnancies and in vivo controls. In contrast, AI-generated fetuses had an average 5mC content of 29%, with a range of 21–34%, and IVF fetuses had an average 5mC content of 27%, with a range of 26–30% (Table 2). The average 5mC content of cloned fetuses from interrupted pregnancies was 23% and ranged from 10% to 28%. The genomic 5mC content of both aborted cloned fetuses (P < 0.001) and cloned fetuses (P < 0.005) was significantly different from the in vivo fetuses. There was no significant difference (P > 0.05) between fetuses generated by AI and IVF. Adult cloned animals had no significant differences (P > 0.05) with respect to genomic 5mC content when compared with AI-derived adult animals. Adult clones had 32% average genomic methylation levels, whereas AI-derived adults exhibited 29%. Methylation-sensitive restriction digests (with isochizomers HpaII/MspI) of genomic DNA from adult clones and in vivo-derived adult animals showed indistinguishable genomic methylation patterns between these two groups (data not shown). Compared with other nucleosides, methylated cytosine was the only nucleoside that was significantly different between cloned and AI-generated fetuses. The percentages of G, A, T, and C in fetal and adult genomes were calculated by the same procedure as 5mC. Nucleoside concentrations (nM) were divided by those of other nucleosides in the genome C/(C + A + T + G), G/(C + A + T + G), A/(C + A + T + G) and T/(C + A + T + G). Percentages of each nucleoside were nearly identical for aborted cloned fetuses, cloned fetuses from ongoing pregnancies, in vivo-derived fetuses, adult clones, and AI-derived adult animals and were not significantly different (P > 0.10) (Table 3). This analysis was also done to address the integrity of DNA samples from some of the aborted fetuses. HPLC clearly demonstrated that there was no detectable loss of nucleosides in DNA samples of fetal or adult origin, which presented similar nucleoside content. Analysis of the unmethylated cytosine content relative to the total cytosine content of fetal genomes also confirmed that, in the aborted cloned fetuses, unmethylated cytosines composed 92% of the cytosine content of the genome, whereas in cloned fetuses from interrupted pregnancies and in vivo fetuses, unmethylated cytosines composed 76% and 69% of the genome, respectively. Unmethylated cytosine levels were also significantly different between in vivo-derived fetuses and aborted cloned fetuses (P < 0.001) and between in vivo fetuses and cloned fetuses from interrupted pregnancies (P < 0.005).
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our findings indicate that NT-generated fetuses have overall decreased levels of methylated cytosines, suggesting a hypomethylated genome when compared with AI-produced fetuses. It is noteworthy to mention that all aborted cloned fetuses survived well beyond the stage where somatic genomic methylation levels are established in normal fetuses. De novo methylation patterns are determined in bovine embryos at the 8–16-cell stage [21]. In mice, the typical genome-wide somatic methylation levels are established after implantation [1, 10], and overall methylation levels are conserved during subsequent organogenesis and differentiation [34]. HPLC analysis performed in this study demonstrated that the overall 5-methylcytosine content of the genome is identical (29%) between 60-day-old AI bovine fetuses and 18–24-mo-old adults. Thus, it is unlikely that global, genome-wide methylation changes arise beyond 60 days. Ongoing 5-methylcytosine changes during further development and differentiation would certainly be loci- or tissue-specific in nature. Age-matched in vivo controls beyond 60 days were not included in this study because overall global methylation levels did not differ between 60-day-old fetuses and in vivo adults.
The extent of decreased genomic methylation varied widely between individual cloned fetuses (0–24%) that aborted spontaneously. It is noteworthy to mention that genomic methylation variation in aborted cloned fetuses defined two distinct categories within this group. Three fetuses, aged 30, 180, and 210 days, exhibited 21%, 24%, and 24% of 5mC content, respectively, whereas the majority of the aborted cloned fetuses presented no detectable levels of genomic cytosine methylation. There was no correlation of these subgroups with donor cell lines or transgenic status. Therefore, despite the prevalence of extensive genomic hypomethylation in aborted cloned fetuses, it is unlikely that fetal lethality and pregnancy failure following nuclear transfer are determined by a single mechanism affecting all individuals. Less variation was observed for Day 60 cloned fetuses whose gestations were apparently proceeding normally, and all fetuses had some detectable level of methylation. These fetuses were characterized by an overall higher level of methylated cytosines when compared with the NT-generated fetuses that aborted but had less than in vivo controls. Less variation was observed between individual fetuses produced by AI. The exact cause for individual variation remains unknown but has also been reported in mice cloned from ES cells [8] and pronuclear exchange [35] as well as cloned bovine blastocysts [20]. It is unlikely that this is a donor cell type-related effect because striking methylation differences were identified between ES cell subclones [8]. We analyzed the DNA methylation status of two genital ridge- and three skin-derived donor cell lines prior to nuclear transfer by hybridization to an interspersed repetitive element. DNA samples from donor cell lines and an in vivo fetal skin control were digested with methylation-sensitive restriction endonucleases and further hybridized with a probe derived from a bovine SINE element (GenBank accession no. AF 181665). Southern blotting results were unable to identify major methylation differences between donor cell lines and the in vivo control (data not shown), as also reported by Kang et al. [20].
The cell types and transgenic status of donor nuclei in the present study and corresponding methylation levels of each fetus and adult are summarized in Table 2. There was no correlation between cell type/origin and abnormal methylation levels because overall hypomethylated genomes were detected in cloned fetuses generated from cumulus-, genital ridge-, and skin-derived cell lines. Similarly, adult clones generated from all three cell types showed global methylation levels comparable with in vivo-derived adult animals. The transgenic status of donor nuclei also did not correlate with genomic methylation patterns detected by HPLC analysis. Thus, it is likely that a common epigenetic reprogramming pathway occurs in nuclear transfer animals irrespective of donor nuclei origin.
Individual variations of methylation levels could be attributed to in vitro procedures, as recently suggested [35]. The overall effects of in vitro culture on genomic methylation levels were addressed in the present study by HPLC analysis of IVF fetuses. No significant differences in genomic cytosine methylation levels were identified between IVF and in vivo-derived fetuses (Table 1). Other studies [4, 22] identified major epigenetic changes in embryos derived from NT but not in those generated by IVF, implying that the effects of in vitro culture on epigenetic reprogramming are much less dramatic than the effects of nuclear transfer. Additionally, cloned fetuses from ongoing pregnancies exhibited various global methylation levels, ranging from 10% to 28% (Table 1), despite the fact that all these fetuses were derived solely from cultured cell lines. It is also clear that in vitro culture of donor nuclei does not prevent adequate epigenetic reprogramming from occurring in a subset of clones, as revealed by similar 5-methylcytosine content between cloned adults generated from cultured cell lines and in vivo-derived adults (Tables 1 and 3).
Cloned fetuses from phenotypically normal, ongoing pregnancies also exhibited hypomethylated genomes when compared with in vivo fetuses of the same stage (2 mo). Although the precise causal mechanisms for these differences remain unclear, it is possible that de novo methylation, which follows the global demethylation surge [34, 36], cannot be fully reestablished in many cloned embryos at postimplantation stages. Mutations in the de novo methyltransferases DNMT 3a and b interestingly result in similar findings as those reported in this study, such as genomic demethylation and perinatal or late development lethality [3]. Recent findings in cloned bovine embryos detected heterogeneous and aberrant de novo methylation following nuclear transfer [21]. Precocious de novo methylation was detected in 50% of early embryos, whereas an opposite methylation-specific staining pattern was detected in the other half of the embryos.
Despite considerable evidence for DNA methylation as the imprinting mark, little is known about the mechanisms that establish genome-wide methylation patterns. Following fertilization, a global demethylation event erases the majority of DNA methylation marks except those of imprinted loci. This is followed by a de novo methylation surge around the time of implantation, which reestablishes overall genomic methylation levels [34, 36]. What are the signals or epigenetic marks that direct the proper reestablishment of methylation patterns across the genome? Epigenetic marks that affect postimplantation de novo methylation of a transgene were identified in the egg cytoplasm of mice [35]. These oocyte-derived marks were established on the transgene upon fertilization. Hypomethylation of the transgene, however, was detected in a subset of pronuclear transplantation embryos, suggesting that embryo manipulation, nuclear transplantation, and culture may have detrimental effects on epigenetic reprogramming.
A complete loss of methylation in the IGF2r differentially methylated region was detected previously in sheep fetuses derived from IVF [18]. This finding was present in 9 out of 12 fetuses diagnosed with large offspring syndrome. This abnormal demethylation may reflect the genome-wide methylation losses identified in our study. Our findings are contrary to global methylation changes detected in cloned bovine blastocysts [20], where 75% of the NT embryos had increased rather than decreased methylation levels. However, these changes were detected in a subset of interspersed repetitive sequences that were used as markers for global DNA methylation status. Nonetheless, because hypomethylation of a satellite repeat was detected in 25% of NT blastocysts in that study, the question arises whether these blastocysts would be more or less likely to reach development to advanced fetal stages or to term. Hypermethylation was also detected in cloned bovine morulae using immunofluorescence [21, 22]. However, the staining pattern of these cloned morulae [21], in the same way as the bisulfite sequencing study [20], was heterogeneous, with many nuclei presenting low, as opposed to high, staining. Monoclonal antibodies directed against 5-methylcytosine are probably limited to detecting methyl groups only in genomic regions in which they are highly concentrated, such as repeats and heterochromatic regions. In addition, this study focused on interphase nuclei as opposed to cells undergoing all cell cycle stages. A simple but crucial difference between the bovine studies so far is that all the global methylation analyses have been done in preimplantation embryos, as opposed to fetuses. In plants, the activation of a defense mechanism, which detects changes in chromatin structure due to global hypomethylation, induces hypermethylation of genomic regions [37]. Although no such mechanism has been described in mammals, hypermethylation of tumor supressor genes, along with global hypomethylation of the genome [38–41], are common features of several types of cancer.
A recent study in mice has linked fetal hypomethylated genomes to perinatal death [42]. Deletions of the Lsh gene, a member of the SNF2 family, results in substantial loss of genomic methylation, comprising repetitive elements as well as single-copy genes. Surprisingly, despite the extensive loss of global methylation levels (13–32% compared with 60–70% in control samples), these Lsh-/- mice developed to term, dying shortly after birth. Thus, abnormal methylation of the genome does not fully impair embryogenesis in the mouse. It remains to be established to what extent the Lsh gene is involved in the high incidence of perinatal death detected in cloned animals across several species.
The finding that adult clones do not differ significantly from in vivo-derived animals, which contrasts with the large variability within cloned fetuses, suggests that there is an epigenetic reprogramming threshold for proper development. It is also feasible that individual cloned fetuses surviving into adulthood have the ability to overcome epigenetic challenges determined by their somatic cell origin. Nonetheless, the data presented here demonstrate that adequate epigenetic reprogramming is a highly variable process in nuclear transfer organisms and may be achieved in only a few nuclear transfer individuals. The complete understanding of how this threshold is achieved in some animals but not in others may provide pathways to increase the efficiency of cloning. It is unlikely that the dramatic fetal and placental abnormalities associated with nuclear transfer are determined solely by DNA methylation losses. However, a genome-wide hypomethylated state, resulting from a possibly aberrant de novo methylation, might be a critical causal factor for the low efficiency and high mortality rates associated with nuclear transfer in animals.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Kenneth J. Eilertsen, Infigen, Inc., 1825 Infinity Dr., DeForest, WI 53532. FAX: 608 846 0520; e-mail: keilertsen{at}infigen.com
Received: 8 August 2002.
First decision: 3 September 2002.
Accepted: 3 October 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. M. Giraldo, D. A. Hylan, C. B. Ballard, M. N. Purpera, T. D. Vaught, J. W. Lynn, R. A. Godke, and K. R. Bondioli Effect of Epigenetic Modifications of Donor Somatic Cells on the Subsequent Chromatin Remodeling of Cloned Bovine Embryos Biol Reprod, May 1, 2008; 78(5): 832 - 840. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Niemann, X C. Tian, W A. King, and R. S F Lee Epigenetic reprogramming in embryonic and foetal development upon somatic cell nuclear transfer cloning Reproduction, February 1, 2008; 135(2): 151 - 163. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Hiendleder, M. Wirtz, C. Mund, M. Klempt, H.-D. Reichenbach, M. Stojkovic, M. Weppert, H. Wenigerkind, M. Elmlinger, F. Lyko, et al. Tissue-Specific Effects of In Vitro Fertilization Procedures on Genomic Cytosine Methylation Levels in Overgrown and Normal Sized Bovine Fetuses Biol Reprod, July 1, 2006; 75(1): 17 - 23. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Constant, M. Guillomot, Y. Heyman, X. Vignon, P. Laigre, J.L. Servely, J.P. Renard, and P. Chavatte-Palmer Large Offspring or Large Placenta Syndrome? Morphometric Analysis of Late Gestation Bovine Placentomes from Somatic Nuclear Transfer Pregnancies Complicated by Hydrallantois Biol Reprod, July 1, 2006; 75(1): 122 - 130. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Hiendleder, C. Mund, H.-D. Reichenbach, H. Wenigerkind, G. Brem, V. Zakhartchenko, F. Lyko, and E. Wolf Tissue-Specific Elevated Genomic Cytosine Methylation Levels Are Associated with an Overgrowth Phenotype of Bovine Fetuses Derived by In Vitro Techniques Biol Reprod, July 1, 2004; 71(1): 217 - 223. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. E. Farin, P. W. Farin, and J. A. Piedrahita Development of fetuses from in vitro-produced and cloned bovine embryos J Anim Sci, January 1, 2004; 82(13_suppl): E53 - 62. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. S. Archer, S. Dindot, T. H. Friend, S. Walker, G. Zaunbrecher, B. Lawhorn, and J. A. Piedrahita Hierarchical Phenotypic and Epigenetic Variation in Cloned Swine Biol Reprod, August 1, 2003; 69(2): 430 - 436. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
