| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Testis |
Department of Cellular and Structural Biology,3 University of Texas Health Science Center at San Antonio, San Antonio, Texas 78229
Department of Biochemistry and Molecular Biology,4
Center for Mammalian Genetics,5
Department of Pediatrics,6 University of Florida School of Medicine, Gainesville, Florida 32611
Department of Biology,7 University of Texas at San Antonio, San Antonio, Texas 78249
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
gametogenesis, gene regulation, spermatogenesis, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In mammals, DNA methylation is typically found only on cytosine residues in the context of a CpG dinucleotide. DNA methylation has been shown to be essential for normal development [7, 8] and has been characterized as an epigenetic mechanism that contributes to the regulation of tissue-specific transcription [9], X-chromosome inactivation [10], genomic imprinting [11], silencing of parasitic sequences [12], cancer progression [13, 14], and aging [15– 17]. Methylation of DNA in the 5' portion of tissue-specific genes generally predicts transcriptional silencing and a repressed chromatin structure [18]. To the extent that DNA methylation is involved in transcriptional repression, demethylation must be required to facilitate transcriptional activation.
Following global changes in DNA methylation associated with early embryonic development, gene-specific changes are seen as specific cell lineages begin to differentiate [19, 20]. In fetal prospermatogonia in the mouse, demethylation of a small number of CpG dinucleotides in methyl-sensitive restriction sites was previously observed in certain genes that are subsequently expressed at later stages of spermatogenesis [21, 22]. However, these studies did not reveal the developmental dynamics of the formation of a demethylation domain. Our hypothesis is that a domain of demethylation develops over the promoter of the phosphoglycerate kinase (Pgk2) gene before transcriptional activation of this gene, that this initiates a cascade of molecular events required to activate transcription, and that this event is signaled by the Pgk2 promoter. The advent of bisulfite genomic sequencing as a tool to examine DNA methylation at all CpG dinucleotides [23], along with our ability to recover highly enriched populations of spermatogenic cell types from fetal, neonatal, prepuberal, and adult mice [24–26], has afforded us the ability to test this hypothesis by examining the ontogeny of the demethylation domain and its timing relative to the occurrence of other molecular events associated with tissue-specific activation of transcription of the Pgk2 gene.
The tissue-specific, autosomal Pgk2 gene is actively transcribed exclusively in meiotic spermatocytes and postmeiotic round spermatids in eutherian mammals [24]. To examine the timing and kinetics of gene-specific demethylation during the development of spermatogenic cells, we used bisulfite genomic sequencing to examine 23 CpG dinucleotides within the endogenous mouse Pgk2 locus in germ cells from male mice at fetal, neonatal, prepuberal, and adult stages. To delineate the gene-specific signal responsible for regulating this process in the Pgk2 promoter, we assessed DNA methylation at up to 46 CpG dinucleotides in transgenes consisting of portions of the human PGK2 promoter ligated to the CAT reporter gene in spermatogenic cell types recovered from transgenic mice at stages before, during, and following active transcription of the Pgk2 gene. Our results suggest that an active, gene-specific process leads to the development of a demethylation domain over the 5' half of the endogenous mouse intronless Pgk2 gene and that this precedes a subsequent cascade of events required for tissue- and stage-specific transcriptional activation during spermatogenesis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Enrichment of spermatogenic cell types was performed on a Sta Put 2– 4% BSA gradient at unit gravity as previously described [24–26]. CD-1 mice (Charles River Laboratories, Wilmington, MA) were used in all cases. Type T1 prospermatogonia were isolated from testes of fetuses at 15.5 and 18.5 days postcoitum (dpc), and type T2 prospermatogonia were isolated from testes of newborn mice at 21.5 dpc. Primitive type A spermatogonia were isolated from testes of prepuberal mice at 6 days postpartum (dpp), and pachytene spermatocytes and round spermatids were isolated from testes of adult (60 dpp) mice. Testicular and vas deferens spermatozoa were isolated from the testes of adult CD-1 mice by dissociation of tissue and sonication to lyse all other cells [24].
Germ cell-type morphology viewed under phase optics was used to determine the purities of each preparation, which were 
95% for spermatocytes, spermatids, and spermatozoa and 85% for prospermatogonia and spermatogonia. Genomic DNA was recovered from each population and from control somatic liver and tail tissues [4, 27]. The Institutional Animal Care and Use Committee of the University of Texas at San Antonio approved all of the animal experimentation protocols.
Transgenic Mice
515PGK2/CAT, 188PGK2/CAT, and CAT-only transgenic mice were prepared as previously described [4, 28]. Briefly, the CAT reporter gene was ligated downstream of 515 base pairs (bp), 188 bp, or 0 bp of the human PGK2 promoter, respectively. Transgenes were introduced into early embryos by pronuclear injection.
Bisulfite Mutagenesis
Bisulfite treatment was used to modify unmethylated cytosines in DNA as previously described [23]. Briefly, 4 µg of genomic DNA was denatured and incubated for 18 h at 50°C in the presence of 2 M sodium bisulfite and 1 mM hydroquinone (pH 5). Samples were then purified using a Wizard Plus minipreps kit (Promega, Madison, WI), eluted in 40 µl water, denatured by incubation with 5 µl 3 M NaOH for 15 min at 37°C, and precipitated. Pelleted DNA was washed once with 70% ethanol, dried, and resolubilized in 50 µl water.
PCR Amplification, TA Cloning, and Sequencing of Bisulfite-Treated DNA
Seven different amplimers were produced from the endogenous Pgk2 gene using seven different sets of polymerase chain reaction (PCR) primers, while up to five different amplimers were produced from PGK2/CAT or CAT-only transgenes. Primer sequences are listed in Table 1. PCR was performed in a reaction mix containing 500 ng bisulfite-converted genomic DNA, 10 mM Gene Amp PCR buffer II, 4 mM MgCl2, 1.5 mM dNTPs, 1.25 µM of each primer, and 1.25 U of Taq Gold (Perkin-Elmer, Norwalk, CT). Amplification conditions were 94°C for 12 min followed by 35 cycles of 94°C for 1 min, 52–60°C for 45 sec, and 72°C for 1 min. PCR products were cloned into the Topo T/A cloning vector (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. Ten bacterial colonies representing unique copies of each amplimer were picked and cultured. Plasmid DNA was isolated from each using the Wizard SV minipreps kit (Promega) and on an ABI 3100 Avant sequencer (Applied Biosystems, Foster City, CA) following manufacturer's instructions. The methylation status of each CpG dinucleotide was determined by comparison of sequences from amplimers of bisulfite-converted and untreated DNA, respectively.
|
Statistical Analysis
Statistical evaluation for comparison of methylation levels between cell types was performed using Fisher exact test for promoter regions, and by chi-square analysis for longer sequences.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Tissue-specific genes are typically methylated and maintain a condensed chromatin structure in cell types in which they are not expressed [9, 18]. The Pgk2 gene is transcriptionally repressed and configured in condensed chromatin in somatic cells [24, 29, 30]. Our analysis of DNA methylation at the endogenous Pgk2 gene in liver and tail showed predominant methylation throughout this locus in both tissues (Fig. 1). The overall frequency of methylation within the Pgk2 gene was 83% in tail tissue and 73% in liver tissue (Table 2). In contrast, the Pgk2 gene is actively transcribed in meiotic spermatocytes and postmeiotic round spermatids and was previously shown to be in a decondensed chromatin structure in these cells [30]. We found a domain of demethylation centered over the Pgk2 promoter and extending approximately 1 kilobase (kb) in each direction in these expressing cell types (note that the upstream portion of this domain consists of an 825-bp region devoid of CpGs) (Fig. 2, E and F). Within this domain, a total of 16 CpGs became hypomethylated to levels of 
25% methylation in spermatocytes and spermatids (Table 2). Together with the region lacking CpG dinucleotides, a demethylation domain of 2.2 kb is formed in expressing cells.
|
|
|
The Pgk2 Demethylation Domain Develops Gradually During Spermatogenesis
Our ability to isolate populations of specific spermatogenic cell types during fetal, neonatal, and prepuberal stages allowed us to examine the ontogeny of the Pgk2 demethylation domain. In type T1 fetal prospermatogonia at 15.5 dpc (Fig. 2A), the Pgk2 locus is 79% methylated throughout. This hypermethylated status mirrors that of somatic cells (Table 2). However, by just 3 days later, in type T1 prospermatogonia at 18.5 dpc, we observed the beginning of a demethylation event in the form of a statistically significant reduction in the frequency of methylation (80% down to 45%) at CpGs 5–8, which reside in the enhancer + core promoter region of the Pgk2 gene (Fig. 2B) (P = 0.001).
In type T2 prospermatogonia at 21.5 dpc (1–2 days after birth), the Pgk2 enhancer + core promoter region has become almost completely demethylated, and hypomethylation has spread to at least 330 bp downstream of the transcription start site (Fig. 2C). By 6 dpp, male germ cells have differentiated into primitive type A spermatogonia and have completed their migration to the basal aspect of the developing seminiferous epithelium, where a subset undertake the role of spermatogenic stem cells [31]. The extent of the demethylation domain has, by this stage, reached a maximum—extending approximately 1 kb in each direction from the enhancer + core promoter region (Fig. 2D). The Pgk2 demethylation domain is bordered by sites 4 and 21, which remained constitutively hypermethylated in all cell types analyzed, spermatogenic and somatic alike (Figs. 1 and 2, A–H).
Remethylation of the Pgk2 Gene Follows Cessation of Expression
Transcription of the Pgk2 gene ceases midway through spermiogenesis, but the demethylation domain persists in testicular spermatozoa (Fig. 2G). However, by the time sperm have traversed the epididymis and collected in the vas deferens, the Pgk2 demethylation domain has become largely remethylated, especially in the 3' portion of the domain (sites 14–20; Fig. 2H). This confirms a surprising de novo methylation activity that is able to function within the context of highly condensed, protamine-rich chromatin [22].
The Pgk2 Promoter Signals Demethylation in Transgenic Mice
Our results show that demethylation is initiated over the promoter of the endogenous Pgk2 gene. This correlates with previous results from studies of transgenes showing that a 188-bp fragment of the human PGK2 promoter is sufficient to direct testis-specific demethylation in cis at two CpG sites within an adjacent CAT reporter sequence in vivo [4]. We have now examined DNA methylation at 46 separate CpG dinucleotides in PGK2/CAT transgenes in spermatogenic and somatic tissues from transgenic mice. The results reveal dramatic differences in the extent of transgene methylation in different cell types. Thus, transgenes with either 515 bp (= core promoter plus upstream enhancer) or 188 bp (= core promoter only) of PGK2 sequence showed averages of 85% and 88% methylation, respectively, at all CpG dinucleotides in liver tissue (Fig. 3, A and E), compared with averages of only 2% and 14% methylation, respectively, in pachytene spermatocytes (Fig. 3, C and F) and 4% and 7%, respectively, in round spermatids (Fig. 3, D and G) (summarized in Table 2).
|
A similar examination of 35 CpG sites within the CAT reporter sequence introduced as a transgene with no adjoining PGK2 promoter sequence (CAT only) also showed extensive methylation overall in somatic cells (Fig. 3H). However, in testis tissue from these same mice, the frequency of DNA methylation was generally more variable and was reduced to an overall average of 42% (Fig. 3I and Table 2). Importantly, while there may indeed be an incompletely penetrant, generic demethylation activity in the testis, the extent of demethylation seen in the CAT coding sequence when ligated to the PGK2 promoter was nearly absolute (24% methylation), and significantly greater than that seen in the CAT-only transgene (42% methylation) (P < 0.001). These results confirm that a cis-acting signal resides within the PGK2 core promoter that is capable of directing demethylation of adjacent sequences in cis in a tissue-, cell type-, and developmental stage-specific manner. Importantly, this signal has been shown to function autonomously in transgenes integrated into different genomic sites and therefore appears to function independently of any other regional or global influences.
To compare the ontogeny of demethylation of the PGK2/CAT transgene to that of the endogenous Pgk2 gene, we analyzed the methylation status of the 515 PGK2/CAT transgene in type T2 prospermatogonia at 21.5 dpc (Fig. 3B). We found that sites 1–11, which reside in the 5' portion of the transgene, are only 39% methylated in prospermatogonia at 21.5 dpc, whereas sites 12–28, immediately downstream from this region, are 76% methylated. Thus, the developmental pattern of demethylation in the transgene appears to mimic that in the endogenous gene in that it begins over the promoter region and then extends in cis.
Developmental Stage-Specific Demethylation Signaled by the Pgk2 Promoter Differs from Constitutive Hypomethylation Signaled by a CpG Island
Interestingly, CpG sites 29–38 in the PGK2/CAT transgene constitute a weak CpG island (224 bp, 47.8% GC, 0.95 observed/expected CpGs) [32] in the 3' portion of the CAT reporter gene (Fig. 3). We found this region to be constitutively hypomethylated in spermatogenic cell types at all developmental stages (Fig. 3, B–D, F, G, I) but hypermethylated in somatic cell types (Fig. 3A). This differs from the demethylation we observed in the 5' half of the transgene, which is developmental stage-specific (occurring in prospermatogonia) and signaled by the Pgk2 promoter. Furthermore, the hypomethylation observed in the PGK2/ CAT CpG island in spermatogenic cells differs from that of typical CpG islands that remain ubiquitously hypomethylated in both somatic and spermatogenic cell types [19] and suggests that the activity responsible for maintaining island hypomethylation is more pronounced in spermatogenic cells than in somatic cells.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
It has been proposed that DNA methylation contributes to the repression of ectopic expression of tissue-specific genes that might otherwise be driven by interactions between ubiquitous transcription factors and core promoter elements [9]. This repression is believed to be mediated in part by methylated DNA-binding proteins that attract histone deacetylases and other chromatin modifying and/or remodeling factors to promote the stabilization of a condensed, transcriptionally inert chromatin structure. Our results demonstrating that the spermatogenesis-specific Pgk2 gene is predominantly methylated in somatic cells support this contention because we previously showed that this gene is not expressed in these cells and is present in a condensed chromatin structure [29, 30]. Our analysis of methylation in the Pgk2 gene in type T1 prospermatogonia at 15.5 dpc indicates this gene is in a similarly repressed configuration in these cells and that demethylation is part of a subsequent derepression mechanism leading to transcriptional activation of this gene later in this same cell lineage.
We previously showed that activation of transcription of the Pgk2 gene is preceded by decondensation of chromatin and the formation of a DNase I hypersensitive site over the Pgk2 promoter in type A and type B spermatogonia at 8 dpp, 2–4 days before the initiation of Pgk2 transcription in primary spermatocytes [29]. Here, we show that this remodeling of chromatin is, in turn, preceded by demethylation of DNA that begins in prenatal prospermatogonia at 18.5 dpc and is essentially complete in primitive type A spermatogonia by 6 dpp. This is consistent with a model in which regulated, gene-specific demethylation initiates a cascade of molecular changes that potentiate transcriptional activation. In particular, this model would predict that demethylation over the 5' portion of the Pgk2 gene leads to a subsequent change in chromatin composition that results in decondensation of chromatin structure. It has been shown that histone deacetylase associates with methyl-DNA binding proteins [33]. Therefore, demethylation of DNA and the resulting dissociation of these methyl-DNA binding proteins along with associated histone deacetylases would favor acetylation of histones and decondensation of chromatin.
Importantly, our results demonstrate that, while demethylation appears to initiate a series of molecular events that predispose transcriptional activation, it is not, by itself, sufficient to initiate transcription. Thus, the Pgk2 demethylation domain is fully developed in spermatogonia by 6–8 dpp, but transcription of this gene does not begin until the formation of primary spermatocytes at 10–12 dpp. Similarly, a PGK2/CAT transgene that carries the Pgk2 core promoter, but not the upstream enhancer elements, develops a demethylation domain in spermatogenic cells but does not initiate transcription. These results suggest that the primary role of demethylation is to signal changes in chromatin structure required to facilitate subsequent binding of additional transcription factors required to initiate transcription (Table 3). This is further exemplified by the lack of active transcription of the Pgk2 gene in testicular spermatozoa despite the persistence of the demethylation domain. The absence of active transcription of the Pgk2 gene in these cells likely reflects the significant chromatin reconfiguration and condensation that occurs during spermiogenesis [34, 35].
|
The timing of the demethylation event we observed is significant in several ways. The development of the Pgk2 demethylation domain is remarkably gradual. However, gradual changes in DNA methylation patterns are not unprecedented. Paternally imprinted genes acquire methylation in an allele-specific manner during spermatogenesis as exemplified by the H19 gene in the mouse. This process is also very gradual, with the paternal allele of H19 acquiring methylation over a 5-day span in prospermatogonia between 13.5–18.5 dpc and the maternal allele requiring >10 days to acquire full methylation, beginning in prospermatogonia at 18.5 dpc but not becoming complete until the formation of primary spermatocytes at 10 dpp [36].
Indeed, analyses of methylation of multiple genes, both imprinted and not imprinted, have indicated that the late fetal and early postnatal stages of development encompass a time of active changes in epigenetic programming of genes in the germ line of both sexes [19, 36, 37]. In the case of the male germ line, the timing of this programming is critical because it precedes the stage at which stem spermatogonia become established, so that a single programming event in the form of a heritable change in a DNA methylation pattern will then be repeatedly propagated to all spermatogenic cells during each subsequent wave of spermatogenesis.
The timing of this event also indicates an active mechanism is involved in establishing the demethylation domain in the Pgk2 gene. Prospermatogonia enter mitotic arrest at about 15.5 dpc and remain in this state until 3–4 days postpartum, when type T2 prospermatogonia undergo a burst of proliferation to seed the seminiferous epithelium [38]. Thus, the demethylation we observe in the 5' half of the Pgk2 gene cannot be due to a passive mechanism in which maintenance methylation fails to occur following DNA replication. Rather, a specific, targeted mechanism is indicated that actively demethylates cytosines within the Pgk2 demethylation domain in prospermatogonia that are not replicating. These results therefore support the notion that DNA demethylase activity exists within these cells [39, 40].
Our data suggest that gene-specific aspects of the core promoter of Pgk2 recruit an active demethylation complex to initiate the demethylation process. We have shown that formation of the demethylation domain is first initiated directly over the promoter of the endogenous Pgk2 gene in the mouse. In addition, our experiments with transgenes show that the human PGK2 core promoter is sufficient to direct the formation of a similar gene-, cell type-, and stage-specific demethylation domain in a reporter gene in cis. Therefore, we hypothesize that one or more elements in the Pgk2 core promoter likely recruit one or more factors that, in turn, attract an active demethylase to initiate bidirectional demethylation.
The endogenous mouse Pgk2 gene contains 23 CpG dinucleotides spread throughout the 5'-flanking region and the intronless coding sequence. However, this gene does not contain a CpG island [34]. The mechanisms that regulate demethylation of this nonisland gene are likely to be distinct from those that regulate demethylation of CpG island-containing genes. Thus, just as different methyl transferases are known to act to add methyl groups to cytosines in island and nonisland sequences, respectively [8, 41], different demethylases and/or cofactors may facilitate demethylation of island and nonisland gene sequences. This is supported by the methylation patterns we observed in different regions of the PGK2/CAT transgenes. The CpG island within the 3' portion of the CAT coding sequence appears to remain constitutively hypomethylated at all stages during male germ cell development, whereas the demethylation domain induced by the Pgk2 promoter in the 5' portion of this transgene undergoes regulated, stage-specific demethylation. The reduced methylation of the CAT-only transgene in the testis appears to reflect a general, albeit partial, nonspecific demethylation activity in this tissue and may be due in part to the presence of the weak CpG island within the CAT reporter sequence. It is also possible that the methylation pattern could be influenced by surrounding sequences or signals at the site of integration of the transgene. However, our data suggest such signals typically function over a relatively limited range.
Finally, remethylation of the Pgk2 gene in vas deferens sperm provides evidence for reprogramming of the Pgk2 gene in conjunction with the transition of the male germline genome from that which directs spermatogenesis before fertilization to that which will contribute to directing embryogenesis following fertilization. The transition between gametogenesis and embryogenesis presents a unique challenge to the germline genome in each sex. Extensive reprogramming of epigenetic regulatory mechanisms is clearly required to facilitate the global changes in gene expression patterns that distinguish spermatogenesis, oogenesis, and embryogenesis, respectively. However, very little is currently known about the mechanisms that regulate this reprogramming.
In summary, we have provided evidence for the regulated development of a gene-, tissue-, and developmental stage-specific demethylation domain that appears to initiate the potentiation of transcriptional activation of the Pgk2 gene in spermatogenic cells. This represents an example of epigenetic regulation of gene expression and establishes a system in which this mechanism and its regulation can be further studied.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence. FAX: 210.458.5658; jmccarrey{at}utsa.edu
Received: 27 February 2004.
First decision: 24 March 2004.
Accepted: 10 May 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Maeda, S. Sato, N. Hattori, S. Tanaka, S. Yagi, and K. Shiota DNA Hypomethylation Circuit of the Mouse Oocyte-Specific Histone H1foo Gene in Female Germ Cell Lineage Biol Reprod, May 1, 2008; 78(5): 816 - 821. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. M. Abhyankar, C. Urekar, and P. P. Reddi A Novel CpG-free Vertebrate Insulator Silences the Testis-specific SP-10 Gene in Somatic Tissues: ROLE FOR TDP-43 IN INSULATOR FUNCTION J. Biol. Chem., December 14, 2007; 282(50): 36143 - 36154. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Suzuki, S. Sato, Y. Arai, T. Shinohara, S. Tanaka, J. M. Greally, N. Hattori, and K. Shiota A new class of tissue-specifically methylated regions involving entire CpG islands in the mouse. Genes Cells, December 1, 2007; 12(12): 1305 - 1314. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Yoshioka, C. B. Geyer, J. L. Hornecker, K. T. Patel, and J. R. McCarrey In Vivo Analysis of Developmentally and Evolutionarily Dynamic Protein-DNA Interactions Regulating Transcription of the Pgk2 Gene during Mammalian Spermatogenesis Mol. Cell. Biol., November 15, 2007; 27(22): 7871 - 7885. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
