| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Center for Reproductive Biology, School of Molecular Biosciences, Washington State University, Pullman Washington 99164-4231
Address all correspondence and requests for reprints to: Michael K. Skinner, Center for Reproductive Biology, School of Molecular Biosciences, Washington State University, Pullman, Washington 99164-4231. E-mail: skinner{at}mail.wsu.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During embryonic development the primordial germ cells migrate down the genital ridge and colonize the indifferent bipotential gonad before sex determination (5, 6). As the primordial germ cells migrate down the genital ridge, their genomic DNA is demethylated such that the germ cells genome before sex determination is not methylated (7). At the onset of sex determination, the germ cell DNA is remethylated in a sex-specific manner (8). Somatic cells in the developing gonads are required for normal germ cell development and DNA methylation (9, 10, 11, 12). Although the majority of the genome is demethylated and remethylated after fertilization and early embryonic development, a small subset of genes are imprinted and transmit a unique methylation pattern to subsequent generations through the male or female germline (13). Modification of the DNA methylation pattern of previously identified imprinted genes has been shown to induce disease states (14). Therefore, alterations in the epigenetic programing (i.e. DNA methylation) of the germline when the cell is establishing its methylation pattern at the onset of sex determination could induce an epigenetic transgenerational disease state.
Endocrine disruptors are a class of environmental and therapeutic compounds that bind to hormone receptors, alter hormone signaling, and influence (i.e. disrupt) the endocrine system (15). These compounds include a wide variety of commonly used substances from pesticides to plastics. Previous studies have demonstrated that exposure to a number of different endocrine disruptors during embryonic or early postnatal development can cause adult onset disease states including reproductive defects (16) and tumor development (17). For example, the therapeutic agent diethylstilbestrol after embryonic exposure causes abnormal female and male reproductive tract development (18, 19), including some F2 generation defects being identified (20). Recently two endocrine disruptors, the fungicide vinclozolin (antiandrogenic) and pesticide methoxychlor (estrogenic, antiestrogenic, and antiandrogenic), have been shown to promote male germ cell defects for multiple generations (F1–F4) after exposure of an F0 gestating mother during the time of gonadal sex determination (1). The ability of an environmental compound (endocrine disruptor) to induce a transgenerational disease state in the progeny of an exposed gestating mother, impacts the potential hazards of these compounds and suggests novel mechanisms in disease etiology.
The phenomena of the fetal basis of adult onset disease is centered on the concept that embryos are more sensitive to environmental factors and alterations in embryonic development can cause the development of a variety of disease states in the adult (21). Examples of factors that can promote adult-onset disease include environmental toxicants (22, 23), abnormal maternal physiology (22), and nutritional support of the embryo (24). Recently observations suggested some of these adult-onset disease states may be transgenerational (1). The only mechanism known to transmit a transgenerational heritable disease state requires the involvement of the germline. The potential that an epigenetic alteration of the germline could be a factor in adult-onset disease has not been previously appreciated. The hypothesis tested in the current study is that embryonic exposure to an environmental factor (endocrine disruptor) at the time of gonadal sex determination induces a permanent alteration in the epigenetic reprogramming (DNA methylation) of the germline that promotes transgenerational disease states. The current study identifies a set of new imprinted-like genes/DNA sequences that are involved in the epigenetic reprogramming of the male germline and appear to be associated with the development of transgenerational epigenetic disease. Although the cause and/or effect of the imprinted-like genes/DNA sequences identified remains to be determined, the novel observation is provided that an environmental compound can induce a transgenerational change in the epigenome of the germline.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
DNA methylation assay
The methylation state of the DNA isolated from inbred strain of Fischer control and vinclozolin F3 generation epididymal sperm was determined using a combination of methylation-sensitive restriction enzymes (MSRE) and PCR procedures previously described (1). Briefly, genomic DNA was isolated from sperm samples using the DNeasy tissue kit (QIAGEN, Valencia, CA). Two micrograms each of control and vinclozolin generation sperm DNA were separately digested with RsaI and either methylation-sensitive enzyme HpaII or insensitive MspI enzymes, followed by PCR with specific primer sets as previously described (supplemental Table S1, published as supplemental data on The Endocrine Society’s Journals Online web site at http://endo.endojournals.org). PCR products were electrophoretically separated on polyacrylamide gels and visualized by ethidium bromide staining. The PCR products (i.e. bands) that were reproducibly absent or present between control and vinclozolin generation sperm DNA were excised from stained gels, reamplified with the same PCR primer set, and then cloned and sequenced. The DNA sequence and chromosomal location was determined using the National Center for Biotechnology Information rat genomic databank BLAST system.
Bisulfite sequencing
Bisulfite sequencing was used to characterize the DNA methylation patterns and changes. Genomic DNA was isolated from SD control and vinclozolin F2 and F3 generation epididymal sperm using the DNeasy tissue kit. Genomic DNA (10 µg) was digested with RsaI and then treated with bisulfite as previously described (1). After purification, the bisulfite-converted DNA was used as a template for PCR. The bisulfite primers for each candidate were designed using the MethPrimer (www.urogene.org/methprimer) system (26) and listed in supplemental Table S1. The sequence-specific primers were generated to amplify the CpG region of interest to characterize the methylation pattern. PCR conditions for amplification were as follows: 30 cycles followed by another 30 cycles with nested primers (30 sec at 95 C, 45 sec at 50 C, and 50 sec at 72 C) and then 5 min at 72 C with modified 1x buffer [10 times: 166 mM (NH4)2.SO4; 670 mM Tris (pH 8.8); 67 mM MgCl2; 100 mM ß-mercaptoethanol]. PCR products were cloned into pGEM-T Easy Vector (Promega, Madison, WI) and sequenced using Big-Dye Terminator (Applied Biosystems, Foster City, CA). Approximately 50 different clones from each PCR product were sequenced to characterize the methylation state of the CpG sites identified. A mixture of a minimum of three different F2 or F3 generation control and vinclozolin animals were used to generate approximately 50 clones per candidate analyzed. Statistical differences between methylation states were determined with a Fisher’s exact test.
Genomic DNA and methylation-sensitive restriction enzyme PCR
Genomic DNA was isolated from the different SD control and vinclozolin F1–F3 generation epididymal sperm samples. A total of 1 µg of genomic DNA was digested using methylation-sensitive HpaII or AciI enzymes (37 C, 6–12 h). The digested DNA was precipitated and dissolved in 30 µl dH2O. Then 3 µl aliquots were analyzed by PCR in 30-µl reactions with specific primers (supplemental Table S1), subjected to 25 cycles, and the PCR products electrophoretically separated on 1.5% agarose. The difference between the control and vinclozolin F1–F3 generation sperm samples was determined for selected genes found to have high optimal digestion capacity.
Microarray analysis
RNA was collected from E13, E14, and E16 testis from control and E16 vinclozolin F1 and F2 generation SD animals as previously described (27, 28). RNA was hybridized to the rat 230A gene chip (Affymetrix, Santa Clara, CA). The Genomics Core in the Center for Reproductive Biology at Washington State University performed the analysis as previously described (27, 28). Briefly, RNA from the cells was reverse transcribed into cDNA and then was transcribed into biotin-labeled RNA. Biotin-labeled RNA was then hybridized to the Affymetrix rat 230A gene chips. Each gene set is composed of 16 pairs of 24-mer oligonucleotides, with one sense strand-specific for the gene and one antisense strand with single-point mutations for use as comparative negative control. Biotinylated RNA was then visualized by labeling with phycoerythrin-coupled avidin. The microarray was scanned on a gene array scanner (Hewlett-Packard Co., Palo Alto, CA). Two microarray chips from two different RNA samples were analyzed for each E13, E14, and E16 control and E16 vinclozolin F1 and F2 generation testis. The microarray data set can be observed online (www.skinner.wsu.edu).
Bioinformatics
The majority of the data analysis was performed using the Affymetrix Microarray Suite software. Most of this study used the comparison analysis software offered and analysis parameters previously described (28). Two repeats for each control and vinclozolin F1 and F2 generation E16 testis was performed and allowed four comparisons in the experiment. Only genes that displayed a consistent expression over all four comparisons and had a relative hybridization intensity of at least 50 were included in the analysis. Basic expression analysis was determined by accessing the Affymetrix database through the Microarray Suite software. A pathway analysis was performed by the Pathway Assist Gene Spring software (Silicon Genetics, Foster City, CA).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The F3 generation samples were used to identify alterations in DNA methylation between control and vinclozolin generation sperm. This analysis assures that any changes observed are due to a transgenerational epigenetic transmission and not due to toxicology that could be present in the F1 generation. The analysis of methylation used a MSRE procedure that involved the use of 10 different sets of PCR primers to consensus CpG methylation sites (MSRE-PCR) (29) as described in Materials and Methods. Sperm DNA from control and vinclozolin F3 generation animals (i.e. Fischer rats) from three different preparations containing a minimum of three different animals were analyzed. A representative gel for one set of PCR primers is shown in Fig. 1
with the three control and vinclozolin samples shown. Differences in the MSRE HpaII (H) digest demonstrated the presence of PCR products (bands) in the vinclozolin samples, but not control samples (Fig. 1). If differences in the PCR products (bands) occurred in a minimum of two of the three replicate animals, they were selected for further analyses. This analysis generated 34 candidates with potential altered DNA methylation between control and vinclozolin sperm samples and were numbered 1–34 (Table 1). The representative differences for candidates 19–22 are shown in Fig. 1 and the other gels and candidates in supplemental Fig. S1, published as supplemental data on The Endocrine Society’s Journals Online web site at http://endo.endojournals.org.
|
|
. The presence of more than one candidate number together (e.g. 2, 3, 24) indicates similar sequences being present in the different candidates. The gene(s) nearest the candidate sequences are listed with GenBank accession numbers. The physical map (i.e. location) of the candidate DNA methylation site is shown as being in the promoter, exon, intron, or distal noncoding region of DNA (Table 1). The sequence was considered in the promoter of a gene if it was within 1000 bp upstream of the transcriptional start site of the gene of interest.
The chromosomal location of all the candidate DNA methylation sequences is shown in Fig. 2. The candidate altered methylation sites are present on various autosomes with no major hot spot regions and none present on the sex chromosomes. The rat genome is not complete so several candidate sequences could not be localized and appear to be on uncharacterized chromosomal regions (Fig. 2). Therefore, the MSRE-PCR screen identified 25 individual candidate DNA sequences with potential altered methylation sites that are different between control and vinclozolin generation F3 sperm DNA samples (Table 1). As discussed below, some of these candidates fall within the promoters, exons, or introns of specific genes, whereas others are noncoding regions distal to a gene and the nearest gene(s) are listed (Table 1).
|
) have specific alterations in DNA methylation requires bisulfite sequencing analysis. Bisulfite converts cytosine residues to thymine residues unless the CpG is methylated at the 5' position on the cytosine, which inhibits the conversion. Specific bisulfite primers were used to amplify the bisulfite converted DNA followed by sequencing of individual clones to assess CpG sites for their methylation state. The bisulfite primers used for each candidate are listed in supplemental Table S1, and the analysis of each of the candidates is described below. Each candidate analysis involved a minimum of three different individual control and vinclozolin F2 and F3 animals with approximately 50 different subcloned DNA samples sequenced. The bisulfite analysis for each individual candidate containing a statistically significant change in methylation is shown in Fig. 3. The chromosomal location of the methylation site and nearest genes are presented, along with the DNA sequence of the candidate site with potential methylated CpG sites underlined, and the asterisks denote specific CpG with statistically different methylation states between control and vinclozolin generation samples. The methylation state for all bisulfite sequence analyses for all the different clones is represented as closed circle for methylated and open circle for unmethylated CpG sites.
|
|
|
|
|
|
Candidates 2, 3, and 24 mapped to 6q12 and are in a noncoding distal region with the nearest genes being leucine rich protein 157 (Lrpprc) and protein phosphatase 1B (Ppm1b) (Fig. 3A). The bisulfite methylation analysis revealed four potential CpG sites and two are hypermethylated (i.e. statistically different) to 44 and 20% (Fig. 4 and supplemental Fig. S2).
|
1E subunit (Cacnale) (supplemental Fig. S3B). The bisulfite methylation analysis revealed six potential CpG sites that did not have any statistical differences in methylation (supplemental Fig. S3).
Candidates 5, 6, and 9 mapped to 4q42 and are in the exon/intron of an uncharacterized gene FLJ22405 protein (Fig. 3B). The bisulfite methylation analysis revealed nine potential CpG sites and two were hypermethylated (i.e. statistically different) to 50 and 35% (Fig. 4 and supplemental Fig. S2).
Candidate 7 mapped to 16q16 and is in a noncoding region distal to ankyrin repeat protein 28 (Ankrd28) and polypeptide N-acetylgalactosaminyl transferase (Galnt12) (supplemental Fig. S3C) and did not have any detectable differences in methylation.
Candidate 8 mapped to 1q42 and is in a noncoding region distal to liprin-1 and Fas (TNFRSF6) associated via death domain (Fadd) (Fig. 3C). The bisulfite methylation analysis revealed five potential CpG sites and one was hypermethylated (i.e. statistically different) to 65% (Fig. 4 and supplemental Fig. S2).
Candidate 10 did not map to a specific chromosome and is uncharacterized (Fig. 2), so methylation with bisulfite analysis was not possible (supplemental Fig. S3D).
Candidate 11 mapped to 4q24 and is in the noncoding region distal to 60S ribosomal protein L7a and uncharacterized RIKEN cDNA 1200009022 (Fig. 3D). The bisulfite methylation analysis revealed three potential CpG sites and two were hypermethyylated (i.e. statistically different) to 49 and 21% (Fig. 4 and supplemental Fig. S2).
Candidate 12 mapped to 14q11 and is in the promoter to the hypothetical gene similar to importin 7, LOC501910 (Fig. 3E). The bisulfite methylation analysis revealed six potential CpG sites and one was hypermethylated to 51% (Fig. 4 and supplemental Fig. S2).
Candidate 13 mapped to 19q11 and is in the intron of Nfix protein (Fig. 3F). The bisulfite methylation analysis revealed 17 potential CpG sites and four were hypermethylated to 50–60% (Fig. 4 and supplemental Fig. S2).
Candidates 14 and 15 mapped to 8q24 and are in the noncoding region distal to leucine rich repeat neuronal 6A (Lrrn6a) (Fig. 3G). The bisulfite methylation analysis revealed four potential CpG sites and two were hypermethylated to 27% (Fig. 4 and supplemental Fig. S2).
Candidate 16 mapped to 17q12 and is in the promoter/exon/intron of optineurin, Fig. 3H). The bisulfite methylation analysis revealed 6 potential CpG sites and one was hypermethylated to 79% (Fig. 4 and supplemental Fig. S2).
Candidates 17 and 18 mapped to 13q24 and are in the bidirectional promoter to nicastrin and coatomer protein complex subunit- (Fig. 3I). The bisulfite methylation analysis revealed 20 potential CpG sites and two were hypermethylated to 55 and 30% (Fig. 4 and supplemental Fig. S2).
Candidate 20 did not map to a chromosome and is uncharacterized (Fig. 2), so methylation with bisulfite analysis was not possible (supplemental Fig. S3E).
Candidates 21 and 32 mapped to 8q23 and are in the exon/intron region of the recently annotated hypothetical gene similar to phosphoglycerate mutase B chain, LOC503205 (supplemental Fig. S3F), with no differences in methylation after bisulfite analysis. The nearest known gene is NCAM1. This potential methylation site is distinct from candidate 1 but in the same hypothetical gene.
Candidate 22 mapped to 8q32 and is in the intron of tgolgin-1 (Fig. 3J). The bisulfite methylation analysis revealed five potential CpG sites and one was hypermethylated to 67% (Fig. 4 and supplemental Fig. S2).
Candidate 23 mapped to 7q22 and is in the noncoding region distal to a hypothetical gene similar to ribosomal protein, LOC503150, and U1 small nuclear ribonucleoprotein subunit (snRP1c) (supplemental Fig. S3G), with no statistical difference in methylation after bisulfite sequencing.
Candidate 25 mapped to 5q24 and is in the noncoding region distal to major urinary protein 4 (Mup4) and hypothetical gene major urinary precursor, LOC502951 (supplemental Fig. S3H), with no statistical difference in methylation detected with bisulfite analysis. However, the candidate 25 sequence is a repetitive sequence present in 12 different locations in chromosome 5q24. Potential methylation changes in the other locations remains to be investigated.
Candidate 26 mapped to 10q26 and is in the promoter of acetyl-co enzyme A carboxylase- (Acaca) (Fig. 3K). The bisulfite methylation analysis revealed five potential CpG sites and two were hypermethylated to 50 and 28% (Fig. 4 and supplemental Fig. S2).
Candidates 27 and 28 mapped to 4q24 and are in the intron of zinc finger protein 212 (Fig. 3L). The bisulfite methylation analysis revealed eight potential CpG sites and two are hypermethylated to 45 and 48% (Fig. 4 and supplemental Fig. S2).
Candidate 29 mapped to 11q11 and is in the noncoding region distal to Runx1 and similar to cell wall protein Awa1p (LOC501772) (Fig. 3M). The bisulfite methylation analysis revealed six potential CpG sites and 1 was hypermethylated to 48% (Fig. 4 and supplemental Fig. S2).
Candidate 30 did not map to a chromosome and is uncharacterized (Fig. 2), so methylation with bisulfite analysis was not possible (supplemental Fig. S3I).
Candidate 31 mapped to 3q23 and is in the noncoding region distal to Wiskott-Aldrich syndrome protein interacting protein (Waspip) and hypothetical protein, LOC499811 (Fig. 3N). The bisulfite methylation analysis revealed four potential CpG sites and one was hypermethylated to 42% (Fig. 4 and supplemental Fig. S2).
Candidate 33 mapped to 6q12 and is in the noncoding region distal to sine oculis homeobox homolog (Six) 3 and predicted Six2 (Fig. 3O). The bisulfite methylation analysis revealed six potential CpG sites and two were hypermethylated to 31 and 57% (Fig. 4 and supplemental Fig. S2).
Candidate 34 mapped to 9q37 and is in the promoter of the GTP-binding protein RAB12 and member of the Ras oncogene family (supplemental Figure S3J). Due to the high GC content of this DNA sequence, bisulfite PCR primers could not be designed to analyze the entire sequence such that methylation analysis was not possible except for a partial region, which showed no change in methylation (supplemental Fig. S3J).
Therefore, of the 25 candidate DNA sequences with altered methylation in vinclozolin generation sperm, 15 were confirmed to have specific hypermethylation as summarized in Fig. 4 and supplemental Fig. S2. Six of the candidates were characterized, but no alterations in DNA methylation were detected (supplemental Fig. S3). Whether methylation occurred in the regions proximal to those analyzed remains to be investigated. These candidates are speculated to have altered DNA methylation due to the procedure used in their isolation but remain to be confirmed with bisulfite analysis. Two candidates, no. 1 (NCAM1/similar to phosphoglycerate mutase) and 34 (RAB12) could not be fully analyzed due to high GC content and will require a different technology to confirm the methylation state of the region of interest.
As a control a known imprinted gene H19 (30) was analyzed to determine whether methylation changes were induced. No change in H19 methylation was detected between control and vinclozolin F2/F3 sperm DNA samples (Fig. 3P).
The control and vinclozolin generation sperm DNA used in the initial MSRE-PCR and bisulfite analysis was from F2 and F3 generation males. Therefore, the epigenetic changes observed in the male germline appear transgenerational and permanently programmed into the sperm DNA. MSRE analysis for specific candidate DNA sequences was developed (Fig. 5). The MSRE analysis involved gene-specific PCR primers to the candidates to obtain a PCR product with the vinclozolin generation sperm DNA vs. control. Analysis of sperm DNA from F1, F2, and F3 control and vinclozolin generation animals demonstrated candidates 1, 14/15, 17/18, 27/28, and 33 all show similar PCR results for F1-F3 sperm (Fig. 5). PCR products were observed in all the vinclozolin generation sperm DNA samples but not in the controls. Therefore, the methylation changes identified are transgenerational and appear imprinted-like in the male germline. Due to the variability in methylation shown in Figs. 3 and 4, many of the candidates and DNA sequences analyzed could not be used to develop a specific MSRE analysis (data not shown). One of the candidates (no. 33) did not have a consensus HpaII digestion site but did reproducibly digest the DNA (Fig. 5), and another candidate (no. 27/28) used an alternate methylation sensitive restriction enzyme AciI. Combined observations demonstrate the epigenetic changes in the male germline identified develop an imprinted-like characteristic and appear transgenerational.
|
, and all others were either not present on the microarray chips or were below the detection limit demonstrating a lack of expression in the E16 testis. The gene expression profile between wild-type E13, E14, and E16 testis is presented (Fig. 6A). The E16 F1 and F2 controls are compared with the E16 F1 and F2 vinclozolin generation expression. A number of the genes had reduced expression in the vinclozolin F1 and F2 generation E16 testis including ankyrin 28, Ncstn, Rab12, and Lrrn6a (Fig. 6). The genes that had increased expression in the vinclozolin F1 and F2 generation E16 testis compared with control were Fadd, Pbm1b, snRP1c, and Waspip (Fig. 6A). Therefore, the effects of the hypermethylation of the genes shown in Figs. 3 and 4 caused both increases and decreases in gene expression in the developing embryonic testis. Whether the changes in gene expression continue to be transgenerational requires future analysis of the F3 generation. The role of these genes in testis development has not been previously investigated.
|
). An internal constitutively expressed gene, ribosomal protein L19, demonstrated the integrity of the RNA and equal loading with control. This decreased expression of NCAM1 was quantitated and a more than 10-fold reduction in expression was observed between control and vinclozolin F3 generation males (Fig. 6C). Combined observations suggest the epigenetic alterations identified in the male germline appear to associate with alterations in gene expression in developing organs. More thorough analysis of the complete effects on gene expression in a variety of different organs is now required. The specific genes associated with the epigenetic changes identified were examined for potential functional relationships. For the genes with known functions or homologs a Pathway Assist bioinformatics program was used to identify functional links between the various genes (supplemental Fig. S4). A number of cellular functions and processes (e.g. proliferation, maturation, motility) are affected by the various genes, but no major clustering was observed among the different genes (supplemental Fig. S4).
A number of the genes associated with the methylation changes identified have a relationship with various disease states (Table 2). Interestingly, all of these genes have been shown to have an epigenetic component to the disease and/or gene identified (Table 2). Therefore, alterations in the epigenetic (DNA methylation) and/or expression of these genes is associated with a number of disease states. In addition to these correlations to previous literature, the vinclozolin generation animals develop a variety of disease states (2). Previously we have demonstrated that vinclozolin (F1–F4) generation animals between 6 and 12 months of age develop a variety of disease states or abnormalities (Table 3) (1, 2). Therefore, the vinclozolin generational animals used develop a variety of diseases (2) and have the altered epigenetic programming of the male germline. Combined observations indicate that the epigenetic alterations in the germline DNA and associated genes identified in the current study are associated with a number of disease states (Tables 2 and 3). The causal relationship of these epigenetic alterations remains to be elucidated.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methylation of DNA is a critical epigenetic factor in the regulation of gene expression (36), mammalian development (37), and disease (e.g. tumorgenesis) (38, 39, 40). Specific DNA methylation patterns can have a role in regulating the ability of promoter regions of genes to respond to transcriptional activator complexes and/or influence chromatin structures that influence gene expression (41). Although the vast majority of genomic DNA methylation is reset after fertilization (8), a small subset of imprinted genes maintain a defined DNA methylation pattern that is transmitted through the male or female germline, resulting in allelic expression differences (42, 43, 44). Imprinted genes are monoallelically expressed in a parent-of-origin manner and recently a bioinformatics approach has identified as many as 600 candidate imprinted genes in the mouse genome (45). Although the specific monoallelic expression pattern was not directly determined for all the genes identified in the current study, the involvement of the paternal allele was determined by examining sperm from multiple generations. None of the rat genes identified or mouse homologs are on the 600 candidate mouse list reported (45), but species differences could be an issue. One of the genes identified Runx1 was on the nonimprinted gene list used (45). The current study presents the novel observation that an environmental factor can reprogram the germline to promote the presence of new imprinted-like genes/DNA sequences. These new imprinted-like genes/DNA sequences acquire an altered DNA methylation pattern that is transferred through the male germline (i.e. paternal allele) to subsequent generations (F1-F3). The current observations suggest an epigenetic reprogramming of the germline is possible. The ability of an environmental factor to alter the transgenerational epigenetic background of an individual and all subsequent progeny has significant impacts on disease etiology.
The caudal epididymal sperm from control and vinclozolin F1-F3 generation males were collected to elucidate potential differences in DNA methylation. The procedure used in the current study involved a MSRE analysis followed by PCR to identify alterations in DNA methylation. This analysis provided 25 different candidate DNA sequences with potential altered methylation. Other procedures involving methylation binding proteins and microarrays (46) have recently been developed. Unfortunately, neither the rat genome or rat gene promoter microarrays are currently available, such that these procedures were not possible in the current study. These procedures may have provided a greater number of candidate sequences, such that the 25 identified should be considered an underestimate of the total potential epigenetic changes possible. Analysis of the 25 candidates demonstrated 15 had confirmed changes in DNA methylation patterns and that these patterns were transgenerational (i.e. imprinted-like). The 10 genes/DNA sequences that were not confirmed are speculated to be imprinted-like due to the procedure involved in their isolation; however, further analysis of flanking DNA sequences is required to identify the specific methylation changes. A number of the altered DNA methylation sites were in noncoding regions distal to the nearest genes. The speculation is that these sites may be involved in chromatin structural changes (e.g. positioning) that influence the distal genes. The NCAM1 expression is an example of this potential regulation. Further analysis of these DNA methylation sites is now required to determine their functional roles.
The expression analysis of several of the genes associated with the candidate methylation sites demonstrated altered gene expression in the F1 and F2 vinclozolin generations embryonic testis. Analysis of one brain-specific gene, NCAM1, demonstrated a dramatic reduction in NCAM1 expression in the adult brain of vinclozolin generation males. Therefore, the impact of the transgenerational epigenetic imprinting on the male germline is an influence on the gene expression of relevant genes in specific organs. Further investigation of the effects of the transgenerational epigenetic phenotype on the transcriptomes of different organs is now required.
The genes associated with the altered DNA methylation sites range from cell adhesion molecules, ion channels, signaling factors such as phosphatases and GTP binding proteins, transcription factors, translational control factors, and membrane proteins (Table 1). A bioinformatics pathway analysis demonstrated no major cluster or direct relationships between the various genes, but cellular processes such as proliferation, maturation (i.e. differentiation), motility, and assembly (i.e. structural) are affected by the genes identified. All the imprinted-like DNA sequences identified were hypermethylated, compared with control generation sperm DNA. This hypermethylation suggests potential mechanisms involved in the imprinting mechanism, but the molecular aspects of imprinting and the role of gonadal sex determination remain to be elucidated. Several factors may contribute to the molecular mechanisms of the altered methylation (i.e. imprinting) observed. The paternal transmission of the phenotype (1, 2) transgenerationally suggests a role of the Y chromosome and/or paternal imprinted genes. Negligible information is available on the rat Y chromosome such that future studies will be required to examine the influence Y chromosome epigenetics may have on the phenotype and altered methylation of the germline. The ability of paternal imprinting to influence the maternal allele also needs to be considered. Because the maternal allele appears to be ignored and/or modified in the paternal transmission transgenerationally, unique imprinting mechanisms are suggested. The potential involvement of paramutation mechanisms also need to be considered in allelic communication (47). Independent of the mechanism involved, the current study demonstrates the ability to alter the germline epigenome that then can transgenerationally influence disease development. Future studies will need to be focused on this mechanism.
A number of human diseases are the result of abnormal epigenetic (i.e. DNA methylation) programming including Angelman’s, Beckwin-Wiedemann, and Prader-Willi syndromes (48, 49, 50). Alterations in the DNA methylation patterns of imprinted genes have also been shown to promote the development of disease (45). Potential epigenetic abnormalities in children from in vitro fertilization (i.e. intracytoplasmic sperm injection) have been identified (51). A study of monozygotic twins suggested environmental epigenetic effects on disease (52). Therefore, numerous studies and clinical conditions suggest epigenetics may be a critical factor in disease etiology.
The previous studies have demonstrated that an environmental factor (endocrine disruptors) can epigenetically reprogram the germline to promote transgenerational disease (1, 2). The current study identified the induction of imprinted-like genes/DNA sequences in the male germline transgenerationally. A number of the associated genes have been shown to be correlated to known disease (Table 2) including Alzheimers (Nicastrin and NCAM1) (53, 54, 55, 56, 57, 58), polymodal nociceptor (Lrppc) (59), bare lymphocyte syndrome (Ankrd 28) (60, 61), hyperglycemia (cacnale) (62, 63, 64, 65), open angle glaucoma (optineurin) (66, 67), acute myloid leukemia (Runx1) (68, 69, 70, 71), Wiskott-Aldrich syndrome (Waspip) (72), and holoprogencephaly (Six3) (73, 74, 75). Therefore, a number of the epigenetic changes identified were associated with genes previously shown to have epigenetic links to human disease. As an example, NCAM1 is critical in neural tube and brain development, and abnormalities in NCAM1 cause a number of brain diseases and abnormalities, including Alzheimer’s (53, 54, 55, 56, 57, 58). Observations demonstrate a dramatic reduction in NCAM expression transgenerationally in the vinclozolin generation adult males. In addition, nicastrin interacts with the 
-secretase complex and is critical for normal secretase membrane function (76). The inhibition of nicastrin alters -secretase cleavage activity and has been linked as a factor in Alzheimer’s disease (77). Epidemiological studies have indicated that Alzheimer’s has a potential paternal transmission and paternal age is a risk factor for this disease (78). Combined observations suggest that an environmental factor can induce an imprinted-like epigenetic change in two different genes associated with brain disease and Alzheimer’s that transmit their epigenetics through the male germline (paternal allele) as a potential factor in the disease. Although the current study identifies a number of imprinted-like genes/DNA sequences associated with disease, future studies are required to determine the cause and/or effect of these alterations in the epigenome on the disease etiology.
Previously we demonstrated that embryonic exposure to vinclozolin during sex determination promotes the development of adult disease transgenerationally (1, 2). These transgenerational diseases include male fertility abnormalities, tumor development, kidney disease, prostate disease, and immune abnormalities (2). The vinclozolin generation animals used in the current study to identify the epigenetic imprinting in the sperm DNA develop disease states at approximately an 85% frequency (2). Therefore, the imprinted-like genes/DNA sequences identified are associated with these disease states, but a direct causal link or correlation remains to be elucidated. Further investigation of these imprinted-like genes/DNA sequences and correlation to various diseases is speculated to identify novel epigenetic diagnostic and therapeutic targets not previously considered.
The concentration of vinclozolin used in the current study is higher than anticipated in the environment. For vinclozolin the lowest observed adverse effect level recommended for humans is 11 mg/kg·d, but doses at the 1 mg/kg·d have biological effects (79). For rats the lowest observed adverse effect level has been reported as 25 mg/kg·d (80), whereas the no observable and adverse effect level is reported as 6 mg/kg·d (81). The environmental levels of vinclozolin have not been rigorously determined, such that no conclusions regarding the toxicology of this compound can currently be made. Toxicology studies to determine whether environmental levels of the compound can induce these disease states are now needed. The mechanism of vinclozolin actions could involve androgen receptor-mediated events and/or toxicity. The androgen receptor has been shown to be present in the embryonic testis at the time of gonadal sex determination in the germ cells, Sertoli cells, and precursor peritubular cells (82, 83). Although the embryonic gonadal steroid production is minimal at this time, androstenedione is produced by the fetal adrenal. Therefore, endocrine disruptors have the capacity to influence embryonic androgen receptor actions (84, 85). Alternatively, vinclozolin actions could involve toxicologic actions on the developing gonad to subsequently influence germ cell development (81, 86, 87). An experiment comparing the actions of a defined antiandrogenic compound such as flutamide with those of vinclozolin is in progress. Further studies are needed to elucidate the endocrine vs. toxicologic actions of vinclozolin on the embryonic testis.
Although epigenetics is associated with a number of diseases and abnormalities, the current major paradigm for disease etiology involves classic genetics and DNA sequence mutations as a major factor. Clearly regional differences in disease frequencies and environmental influences have suggested additional factors in the development of disease. The observations presented in the current study demonstrate that environmental compounds (endocrine disruptors) can induce a reprogramming in the epigenetic imprinting of the germline. The ability of an environmental factor to promote a permanent epigenetic change in the germline suggests an epigenetic component in disease etiology and a molecular mechanism for the ability of environmental factors to influence disease. This transgenerational epigenetic mutagenesis involves the ability of an environmental factor (e.g. endocrine disruptor) to influence embryonic development at the time of sex determination to epigenetically reprogram the germline through the induction of new imprinted-like genes/DNA sequences that then lead to epigenetic alterations in gene expression to promote disease states transgenerationally. Further analysis of this epigenetic disease etiology will provide novel epigenetic diagnostics and therapeutic targets to advance disease therapy. In addition to disease etiology, this transgenerational epigenetic mutagenesis also impacts basic developmental biology and broader areas of biology such as evolution.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Current address for H.-S.C.: National Research Institutes, Taipei, Taiwan, Republic of China.
Disclosure summary: H.-S.C., M.D.A., S.S.R., and M.K.S. have nothing to declare. None of the authors have a financial conflict of interest with this manuscript.
First Published Online September 14, 2006
Abbreviations: E, Embryonic day; Fadd, Fas associated via death domain; MSRE, methylation-sensitive restriction enzymes; Mup4, major urinary protein 4; NCAM1, neural cell adhesion molecule 1; SD, Sprague-Dawley; Six, sine oculis homeobox homolog; snRP1c, small nuclear ribonucleoprotein subunit; Waspip, Wiskott-Aldrich syndrome protein interacting protein.
Received July 25, 2006.
Accepted for publication August 30, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) denoting nonmethylation and closed circle (
) methylation. Each series of circles represents individual cloned DNA sequences analyzed. The CpG with statistically significant (P < 0.05) altered methylation in the vinclozolin generation sperm DNA is marked with an asterisk (*).
