This page has been archived and is no longer updated
Transposable elements and the epigenetic regulation of the genome
Author: R. Keith Slotkin ,Robert Martienssen
Keywords
Keywords for this Article
Add keywords to your Content
Save
|
Cancel
Share
|
Cancel
Revoke
|
Cancel
Rate & Certify
Rate Me...
Rate Me
!
Comment
Save
|
Cancel
Flag Inappropriate
The Content is
Objectionable
Explicit
Offensive
Inaccurate
Comment
Flag Content
|
Cancel
Delete Content
Reason
Delete
|
Cancel
Close
Full Screen
"Genome sequencing has revealed that transposable elements (TEs) of various classes constitute a large frac- tion of most eukaryotic genomes, including nearly 50% of our own 1 . Active TEs are highly mutagenic, often targeting protein-coding genes for insertion, and also causing chromosome breakage, illegitimate recombina- tion and genome rearrangement. TEs can also influence neighbouring genes by altering splicing and polya- denylation patterns, or by functioning as enhancers or promoters 2 . TEs are often considered as ?selfish? or ?parasitic? ele- ments, because their success (that is, an increase in copy number) is negatively correlated with the fitness of the host. However, most TEs are not actively transposing or duplicating. Although in most cases this is due to muta- tions and deletions that abolish transposition, some full-length autonomous TEs remain intact but silent in host genomes, in the form of cryptic elements. To combat the potentially harmful effects of active TEs, the genome has evolved epigenetic ?defense? mechanisms to suppress their activity. An epigenetically inactive TE retains the coding potential to mobilize itself but does not produce the necessary proteins owing to a repressive chromatin environment. It has long been known that cryptic and non-autonomous TEs accumulate in heterochromatic regions of the genome. However, recent discoveries concerning the molecular mechanisms by which TEs are epigenetically silenced are increasing our understanding of their contribu- tion to the function of heterochromatic regions that flank centromeres and telomeres. It is becoming clear that TEs have become part of the regulatory toolkit of the genome, with important roles in directing gene expression. Recent studies have implicated TEs in imprinting and X-chromosome inactivation, as well as in the developmental regulation of gene expression, a role that was first described by McClintock shortly after her discovery of TEs 3 . Here we bring together the current understanding of TEs as epigenetic regulators of the genome, drawing on studies from yeast, plants and animals. Silencing mechanisms that suppress TEs Genome-wide screens for genes that are required for the regulation of class I TE activity in budding yeast 4 and Class II TE activity in Caenorhabditis elegans 5 (see BOX 1 for an overview of the different types of TE) have uncovered more than 100 genes with various functions. Most of the genes that were identified did not overlap between the two screens, suggesting that many more genes might alter TE activity. Here we focus on epige- netic mechanisms of TE silencing at the transcriptional and post-transcriptional levels, which involve a host of factors that regulate TE mobility and carry out important cellular functions. Post-transcriptional silencing of TEs by RNAi. RNAi is the mechanism by which dsRNA is cleaved by members of the dicer family of proteins into short 21?30-nucleotide small interfering RNAs (siRNAs) that guide RNA- degrading complexes to a complementary transcript (FIG. 1a). Argonaute proteins constitute the catalytic component of the siRNA-guided transcript-cleavage complex, which is known as RISC. RNAi is the main mechanism of TE silencing in C. elegans, in which genes that are required for RNAi are essential for silencing the Cold Spring Harbor Laboratory, 1 Bungtown Road, Cold Spring Harbor, New York 11724, USA. Correspondence to R.M. martiens@cshl.edu doi:10.1038/nrg2072 Transposable elements All mobile DNA segments in the genome, regardless of their mechanism of transposition. Transpose The movement of a genetic element from one location of the genome to another. Autonomous element A transposable element that produces all the proteins that are required for transposition. Cryptic element A transposable element that is epigenetically inactivated for a period of time and has lost the ability to mobilize. Epigenetic A heritable change that is not caused by a genetic mutation. Transposable elements and the epigenetic regulation of the genome R. Keith Slotkin and Robert Martienssen Abstract | Overlapping epigenetic mechanisms have evolved in eukaryotic cells to silence the expression and mobility of transposable elements (TEs). Owing to their ability to recruit the silencing machinery, TEs have served as building blocks for epigenetic phenomena, both at the level of single genes and across larger chromosomal regions. Important progress has been made recently in understanding these silencing mechanisms. In addition, new insights have been gained into how this silencing has been co-opted to serve essential functions in ?host? cells, highlighting the importance of TEs in the epigenetic regulation of the genome. REVIEWS 272 | APRIL 2007 | VOLUME 8 www.nature.com/reviews/genetics AAAAAAA LTR Non-LTR Non-autonomous Autonomous Autonomous helitron Mutation derivative TIRs conserved MITE TIR Transposase TIR TIR TIR Type I ? Retrotransposons Type II ? DNA transposons orf1 orf25? UTR TIR Transposase TIR TIR TIR Replicase Helicase LTR LTR gag pol Non-autonomous element A transposable element that does not produce the proteins required for transposition. These elements are dependent on the proteins produced by autonomous elements of the same element family to transpose. Heterochromatin The portion of interphase chromosomes that remains densely stained and condensed after cell division. Heterochromatic regions are generally gene poor, replicate late, and have a low level of transcription. Imprinting A gene is expressed from only one allele in a diploid organism, depending on the parent of origin. Duplicative transposition A mechanism of transposition that results in a copy of the element at both the excision and acceptor site. This can occur even after excision of the element by the process of gap repair. Tc1 DNA transposon in the germ line 6,7 . In addition, muta- tions in both argonaute- and dicer-family proteins cause the reactivation of TEs in many eukaryotic species. TEs give rise to numerous siRNAs 8 in most species, includ- ing the recently reported TE siRNAs in humans 9 , and TE-siRNA levels are correlated with element activity 7,10 . Although our understanding of how TEs are specifically targeted by the RNAi pathway is incomplete, structural features of TEs might help to distinguish their tran- scripts from those produced by host genes. For example, transcripts containing the terminal inverted repeats (TIRs) of DNA transposons might fold back to produce a hairpin dsRNA structure. Other mechanisms that are responsible for triggering RNAi are explored in BOX 2. Chromatin modifications. A range of chromatin modifi- cations suppress TE transcription, including modifications of histone tails, DNA methylation and alterations in chromatin packing and condensation. Modifications of histone amino (N)-terminal tails alter the binding of pro- tein factors and relay information to transcription factors. Nucleosomes that are associated with TEs are enriched for methylation of histone H3 at lysine 9 (H3K9), which is a signal for transcriptionally repressive and inactive chromatin 11,12 . Mutations in genes required for repres- sive histone tail modifications lead to TE reactivation; for example, mutations in the histone H3K9 methyltrans- ferase gene Suv39 result in modest upregulation of TE transcripts in mouse embryonic stem (ES) cells 12 . Although it is not present in all eukaryotes, DNA methylation on cytosine residues is another important signal that represses TE transcription. In both plants and mammals, cytosine residues can be methylated in a symmetrical context (CpG), and this methylation is copied to the new DNA strand upon DNA replication, providing a mechanism for inheritance of TE silencing. In the mouse, the DNMT1 DNA methyltransferase is responsible for this maintenance of DNA methylation. Elevated intracisternal A-particle (IAP) retrotransposon transcript levels have been reported in Dnmt1-deficient embryos, which die after 9?10 days 13 , but not in undif- ferentiated ES cells 12 . Asymmetrical DNA methylation also targets TEs for silencing in plants and, to a lesser extent, in mammals 14 . However, this methylation must be targeted to the TE de novo with each generation, and does not require partially methylated substrates as a guide. In the mouse, DNMT3 methyltransferase carries out de novo (mostly CpG) DNA methylation, and the non-catalytic homologue DNMT3L is required for IAP retrotransposon methylation and silencing in premeiotic male germ cells 15 . These findings sug- gest that DNA methylation might be required for the epigenetic silencing of TEs at specific stages of mam- malian development (see below). The mechanism that is responsible for targeting de novo and asymmetrical DNA methylation is discussed later with respect to RNA-directed DNA methylation (particularly in the model plant Arabidopsis thaliana). Box 1 | Types and structures of transposable elements The diverse array of transposable elements (TEs) can be classified using several criteria, such as the requirement for a reverse-transcription step to transpose. TEs that require a reverse-transcription step are called retrotransposons, or type I elements. They can be divided into two types, on the basis of the presence or absence of direct repeats at the ends of the element called long terminal repeats (LTRs). LTR retrotransposon proteins pol and gag are closely related to retroviral proteins; however, these elements lack the envelope protein that is required to exit the cell. Retrotransposons undergo duplicative transposition, as their total number increases after each transposition with the potential to expand genomes. For example, one non-LTR retrotransposon family, LINE1, constitutes 17% of the total human genome 1 . Type II TEs, known as DNA transposons, do not require a reverse-transcription step to integrate into the genome. Instead, a transposon-encoded protein called a transposase recognizes the terminal inverted repeats (TIRs) that flank the TE, excises the TE out of the donor position, and then integrates the transposon into the new acceptor site. The gap that is left at the donor position can then be either repaired without element replacement in cut-and-paste transposition, or filled with a copy of the transposon by gap repair. Helitrons are newly identified DNA transposons that duplicate differently, using a rolling-circle mechanism. Autonomous helitrons contain a DNA helicase protein, as well as a replicase protein similar to replicon protein A (RPA). TEs can also be classified by their self-sufficiency. Both retrotransposon and DNA- transposon families have autonomous elements and non-autonomous elements. Non-autonomous elements are often mutated relics of autonomous family members, but sometimes have only limited sequence similarity to their autonomous counterparts. Non- autonomous DNA transposons often consist of a pair of TIRs surrounding non-transposon DNA, which is frequently copied from an unrelated protein-coding gene. Additionally, some non-autonomous elements, called miniature inverted-repeat transposable elements (MITES), consist of two TIRs that are connected in a tail-to-tail orientation. The number of non-autonomous elements in a genome can greatly outnumber the autonomous elements. For example, there are an estimated 1.2 million copies of the Alu repeat in the human genome (reviewed in REF. 122), which is a non-autonomous non-LTR retrotransposon (also known as a short interspersed nuclear element, or SINE). REVIEWS NATURE REVIEWS | GENETICS VOLUME 8 | APRIL 2007 | 273 FOCUS ON EPIGENETICS a RNAi RISC RNA Cleavage b RNAi-mediated chromatin modifications RNA siRNAs RNA RNA RITS siRNA-loaded RITS dsRNA Cleavage Euchromatin ? Active transcription Dicer RNA siRNAs Dicer siRNA-loaded RISC Transcript containing complementary sequence Methylation at histone H3K9 CpG DNA methylation DNA dsRNA formation RdRP Recruitment of: ? Histone methyltransferase ? De novo DNA methyltransferase ? Other chromatin-modifying proteins Heterochromatin ? Reduced transcription Cut-and-paste transposition A mechanism of transposition in which no copy of the element remains at the excision site. Proteins that modify chromatin structure are also involved in TE silencing, and several of these proteins are implicated in chromatin packaging and condensation. In plants, SWI/SNF chromatin-remodelling proteins, which use the energy of ATP hydrolysis to remodel nucleosomes, are required for TE silencing. In A. thaliana, the SWI2/SNF2-like protein DDM1 is specifically required for TE silencing and chromatin condensation 16 . Interestingly, TEs remain activated, or ?preset? (BOX 3), when passed from a ddm1 mutant to a wild-type background 17 , and do not regain DNA and H3K9 methylation, both of which are lost in ddm1 mutants 16 . In the mouse, the ddm1 homologue Lsh1 also remodels chromatin and suppresses TEs, which lose DNA methylation in Lsh1 ?/? embryos 18 . Transcriptional analysis indicates that, although TE transcripts are strongly elevated in ddm1 and Lsh1 mutants, most genes are unaffected, suggesting that the remodelling activity is specifically targeted to TEs 19 . RNAi-mediated chromatin modifications. In a number of eukaryotes, RNAi components are required for chro- matin modifications. Although the maintenance of some repressive chromatin modifications does not require RNAi, it is not known whether these modifications are initially established by RNAi-based mechanisms. It is currently thought that the maintenance of TE silencing at the chromatin level involves redundant functions of both RNAi-dependent and RNAi-independent path- ways, although our understanding of RNAi-mediated chromatin modifications is incomplete. In the fission yeast, Schizosaccharomyces pombe, heterochromatic silencing depends on transcription by RNA Polymerase II and on the RNAi machinery (reviewed in REF. 20). Pericentromeric heterochromatin in S. pombe has provided a model for RNAi-mediated chromatin modification, which parallels TE silencing in using a similar set of siRNA triggers and silencing fac- tors. In S. pombe, the RNAi machinery works on nascent RNAs, which are still bound to the chromosome and are processed co-transcriptionally into siRNAs 21 (FIG. 1b). These siRNAs are used by an argonaute protein (Ago1) to cleave nascent RNAs, and this cleavage recruits the histone H3K9 methyltransferase Clr4 to the chromo- some 22 by an unknown mechanism. RNAi is specifically required for spreading of H3K9 methylation from repeats into reporter genes that are embedded into heterochro- matic transcripts 22 , similar perhaps to the spreading of heterochromatic silencing that is observed in position effect variegation (PEV) in Drosophila melanogaster (see below). In S. pombe, both pericentromeric repeats and Tf2 LTR retrotransposons are under the control of histone modifications, but the role of RNAi in Tf2 silencing is modest by comparison 23 . In the ciliated protozoa Tetrahymena and Paramecium, small RNAs target regions of the genome for chromatin modification, but these target regions subsequently undergo developmentally regulated DNA deletion (reviewed in REF. 24). Argonaute, dicer and histone methyltransferases are all required for this process, in which siRNA from one nucleus are somehow compared with transcripts from another nucleus to select sequences for elimination. In many ways, these internal eliminated sequences (IESs) resemble DNA transposons, but, unlike transposons in other eukaryotes, they are silenced and excised at the same time. This interesting system is likely to yield future insights into the epigenetic regulation of TEs. Figure 1 | Mechanisms of transposable element silencing. a | In the post- transcriptional RNAi pathway, dsRNA is cleaved into small interfering RNAs (siRNAs) by a dicer-family protein. These siRNAs are incorporated into the RISC complex, which contains an argonaute-family protein. The siRNA-loaded RISC complex then cleaves transcripts that are complementary to the siRNA sequence. In this example, dsRNA is initially formed by overlapping antiparallel transcripts. RdRP can additionally produce dsRNA from the cleaved RNA products (BOX 2). b | siRNAs can feed into an alternate pathway by which a different complex containing an argonaute-family protein (called RITS in Schizosaccharomyces pombe) guides the cleavage of nascent transcripts, which are still attached to RNA polymerase II and the DNA strand. Cleavage of this nascent transcript targets this region of chromatin for modification by the recruitment of a histone H3 lysine 9 (H3K9) methyltransferase and other proteins. Modifications of heterochromatin include the methylation of the H3K9 and possibly methylation of the cytosine bases of DNA in organisms in which DNA is methylated. Modifications to transposable element chromatin result in heterochromatin that is condensed and inaccessible to transcription. REVIEWS 274 | APRIL 2007 | VOLUME 8 www.nature.com/reviews/genetics TIR TIR siRNAs RdRP RdRP RdRP TIRTIR Gap repair A process whereby, after excision of a DNA transposon, the element can be copied back into the excision site from the template of the sister chromatid or homologous chromosome. DNA transposon Transposable elements that do not use a reverse-transcription step to integrate copies into the genome. Terminal inverted repeat Repeats that flank most DNA transposons and lie in an inverted orientation. Retrotransposon Transposable elements that use a reverse-transcription step to integrate copies into the genome. Also known as retroposons. Presetting Describes the inheritance of transposable-element- mediated gene control that is independent of the transposable element. Usually confined to the first generation. RNA-directed DNA methylation and TE silenc- ing have been investigated extensively in A. thaliana. The A. thaliana RNAi gene families have duplicated and diversified to participate in post-transcriptional regulation as well as siRNA-mediated chromatin modifications 25 . For example, A. thaliana contains four different dicer proteins, each with specific yet overlap- ping functions. Likewise, two distinct size classes of siRNA have been identified in this species 26 . The smaller class of 21?22-nucleotide siRNAs are involved in post- transcriptional regulation (RNAi and microRNAs), whereas the larger class of 24?26-nucleotide siRNAs are involved in RNA-dependent DNA methylation, and are derived predominantly from TEs and tandem repeats. The DICER-LIKE 3 (DCL3) gene generates the larger 24?26-nucleotide siRNA size class from TEs, and one of the ten argonaute proteins, AGO4, binds a subset of these siRNAs. The siRNA-loaded AGO4 complex seems to have both catalytic and non-catalytic roles in silencing different TEs by targeting asymmetrical DNA methylation 27 . Heterochromatin formation targeted by siRNAs raises an interesting question: how do TEs that are silenced at the chromatin level produce transcripts that are to be cleaved into siRNAs? In A. thaliana, this is accomplished by the plant-specific RNA polymerase IV (REFS 28,29). Loss of Pol IV can transcriptionally reactivate several silenced TEs, as well as isolated solo long terminal repeats (LTRs) 30 . There are two forms of A. thaliana Pol IV (Pol IVa and Pol IVb), which are distinguished by their unique largest subunits. Both forms have a specific role in RNA-directed DNA methylation: Pol IVa in generating siRNAs, and Pol IVb in transferring the siRNA signal to produce the silencing outcome 31,32 . Pol IVb is probably guided to the correct location by its carboxy (C)-terminal domain, which is significantly longer than that of Pol IVa. This C-terminal domain interacts directly with AGO4 (REF. 33) in specific nuclear bodies in an RNA-dependent manner 34 , and might serve as a platform for different pro- tein factors. It is possible that Pol IV-generated ?trigger? transcripts that are produced from full-length or cryptic TE elements are targeted co-transcriptionally as they are in S. pombe. It is also possible that Pol IV acts directly on an RNA template (rather than DNA), generating dsRNA. Mutant screens for genes that are involved in RNA- directed DNA methylation have recovered both Pol IV and a SWI/SNF chromatin-remodelling gene, DRD1 (REF. 35). Along with DDM1, DRD1 might cooperate with Pol IV in transcribing heterochromatic DNA (or even producing dsRNA), and subsequent targeting of these regions for DNA methylation. Although symmetrical DNA methylation is herit- able, asymmetrical DNA methylation must be replaced at each cell division by a targeted mechanism, perhaps involving siRNAs. The A. thaliana Dnmt3 homologue, DRM2, cooperates with another methyltransferase, CMT3, to impose asymmetrical DNA methylation, and is required for RNA-dependent DNA methyla- tion (reviewed in REF. 36). Asymmetrical methylation of TEs in A. thaliana requires DRM2 and CMT3, but transgenes and TEs are only selectively reactivated in Box 2 | The origin of dsRNAs Small RNAs can initiate silencing at the post-transcriptional level or lead to chromatin modifications (FIG. 1). Although the process by which small interfering RNAs (siRNAs) are generated from dsRNA is well understood, the origin of the dsRNA from transposable elements (TEs) remains mysterious. Owing to their inverted-repeat structure (BOX 1), DNA transposons might form a hairpin upon read-through transcription of the entire element from external promoters. Such read-through has been detected in Caenorhabditis elegans for the Tc1 DNA transposon, and it was proposed that this process initiates siRNA production from the Tc1 terminal inverted repeats 7 . Likewise, read-through of a defective inverted-repeat transposon derivative is a dominant repressor of a maize DNA transposon family because it works as an siRNA trigger 123 . By this mechanism, the TE is continuously transcribed while silenced, because the element itself does not initiate transcription. dsRNA can also be generated from two overlapping antiparallel transcripts that base pair together. For example, bidirectional transcription from the LINE1 5? UTR in human cells accounts for siRNA production from this element 9 . Insertion of a small piece of a TE into a bidirectionally transcribed region of the genome might also initiate transposon silencing by producing dsRNA. For example, in Drosophila melanogaster, insertions of TEs into heterochromatic loci result in the Aubergine- dependent germline silencing of homologous TEs elsewhere in the genome 124 . Similarly, in fission yeast, reporter genes that are inserted into heterochromatin are silenced by read-through transcription and RNAi 22 . In plants, C. elegans and fission yeast, dsRNA is also produced by a family of RNA- dependent RNA polymerases (RdRPs), which copy ssRNA templates into dsRNA. RdRP proteins participate in TE-mediated heterochromatin production 25 . How RdRP proteins select transcripts to feed into the RNAi and heterochromatin pathway is not currently understood. It might be that RdRPs are drawn to the RNA template by an siRNA-loaded complex, or by processing of sliced transcripts, suggesting that RdRPs might be involved only in producing secondary siRNAs along the length of the transcript template, and not in the production of the primary siRNA that is needed to initiate silencing. In plants, a recent report suggested that transcripts with two (or more) complementary microRNA sites might preferentially trigger RdRP 125 . (TIR, terminal inverted repeat). REVIEWS NATURE REVIEWS | GENETICS VOLUME 8 | APRIL 2007 | 275 FOCUS ON EPIGENETICS Position effect variegation Unstable gene-expression patterns that are determined by the location of a gene. LTR retrotransposon Long terminal repeat retrotransposons are a type of retrotransposon that has repeats in a direct orientation at either end, in contrast with non-LTR retrotransposons. Transposable element cycling The transformation of a transposable element from an epigenetically inactive state to an active state, and back to an epigenetically inactive state, over the course of development or over multiple generations. Polycomb group A class of proteins, originally described in Drosophila melanogaster, the function of which is to maintain stable and heritable epigenetic repression. non-LTR retrotransposon A retrotransposon that is not flanked by long terminal repeats. these mutants if the symmetrical CpG maintenance methyltransferase MET1 (the homologue of Dnmt1) is still present 37 . This indicates that asymmetrical DNA methylation is much less important than symmetri- cal CpG methylation in maintaining TE silencing 10 . The observation that only a subset of A. thaliana TEs are regulated by RNAi in the presence of MET1 and DDM1 reinforces the importance of chromatin remodelling and CpG methylation in maintaining TE silencing. The mechanisms involved in RNAi-mediated chro- matin modifications are not as well understood in mammals. It is known that the DNA methyltransferase Dnmt3a binds to artificially introduced siRNAs 38 , and artificial siRNA can direct DNA methylation, which is consistent with a requirement for this enzyme downstream of RNAi 36 . Germline silencing and the Piwi/Aubergine pathway. In animals, the germ line is sequestered early in embryogenesis, making this tissue uniquely sensitive to the heritability of TE activity. In D. melanogaster, crosses between males with P DNA transposons and females that are devoid of P elements result in sterility associated with high rates of germline transposition in the progeny, phenotypes that are collectively referred to as hybrid dysgenesis. P elements in the maternal parent suppress hybrid dysgenesis, suggesting the involvement of a cytoplasmic factor, or cytotype, that suppresses the activity of these elements (reviewed in REF. 39). P elements that are inserted near the X-chromosome telomere strongly induce the repressive nature of the maternal cytotype, and this location is a hot spot for P-element transposition. Recently, silencing mediated by these telomeric P elements was found to depend on the Polycomb group gene Enhancer of zeste and the argonaute protein Aubergine 40 . These results suggest that RNAi and heterochromatin might be respon- sible for the maternal repression of subtelomeric P elements. One idea is that heterochromatin-producing non-coding RNAs that are transcribed from telomere- associated repeats might extend into P elements that are inserted within these repeats, resulting in siRNA-mediated trans silencing of active P elements genome-wide 41 . Similar to P elements, non-LTR retrotransposons are also responsible for hybrid dysgenesis in D. melanogaster. Transformation of an I non-LTR retrotransposon into a genome that is devoid of I elements results in an increase in element copy number before the entire fam- ily of elements is eventually silenced. Silencing of the I element can be induced by introducing short pieces of the I element before transformation with the full-length element. This silencing requires transcription (but not translation) and is orientation-independent 42 . Similar to P cytotype, I activity depends on Aubergine 43 , and the repressive signal from specific I elements that are located in heterochromatin is passed from females to their offspring. Therefore, two different heterochro- matic TEs (P and I) produce signals that are transmitted in the maternal cytoplasm to protect the progeny from the detrimental effects of element mobility and hybrid dysgenesis. These signals are likely to be small RNAs, and they are important for genome integrity. Similarly, although they do not contribute to hybrid dysgenesis, Gypsy LTR retrotransposons are regulated in trans by the Flamenco locus, which contains Gypsy and other LTR retrotransposon fragments 44 . Silencing by Flamenco also depends on the argonaute protein Piwi (reviewed in REF. 41). P, I and Gypsy elements are all silenced in the D. melanogaster germ line, precisely where the argo- naute proteins Piwi and Aubergine are expressed. This might account for the activation of several classes of D. melanogaster TEs in cultures of somatic cells that lack piwi and aubergine expression 45 . Consistent with such a role, the Piwi/Aubergine subclade of argonaute- family genes is essential for germline development, not only in Drosophila, but also in mammals and C. elegans 46 . piwi/aubergine mutants have phenotypes that suggest that they are involved in regeneration of germ cells, and a recent report provides evidence that aubergine developmental phenotypes might result from a triggering of the DNA-damage pathway caused by excessive dsDNA breaks 47 , perhaps generated by TEs. Interestingly, the Piwi subclade is not found in plants, which do not sequester germline cells early in development. Piwi proteins bind special small RNAs (termed piRNAs) that are larger (26?30 nucleotides) than most small RNAs 48,49,50 . In mammals, many piRNAs in the male germ line are derived from TEs, but most (80%) are derived from approximately 140 loci that correspond to long non-coding RNAs, which are not known to be TEs 49,50 . piRNA clusters in mice and humans occur in syntenic locations, suggesting a conserved function in germline development; however, they are not related in sequence. These clusters are strand-specific, suggesting that the piRNA precursors are transcribed from only one strand. In D. melanogaster, piRNAs are referred to as repeat-associated siRNAs, or rasiRNAs 51 , because they are generated from repeats and TEs 43 that are clustered in the genome. Drosophila melanogaster rasiRNAs have 3? modifications that differ from other small RNAs 43 , but it is not clear whether piRNAs in other species (such as mammals) carry these same modifications. Aubergine Box 3 | Transposable element cycling versus presetting In her seminal research on transposable elements (TEs), Barbara McClintock described cycling and presetting of the Spm class of DNA transposons. Autonomous TEs cycled from active to inactive epigenetic states and back again in a single generation, or in multiple generations. By contrast, when a non-autonomous TE-generated epiallele responded to the activity of an autonomous element in trans, the epiallele sometimes continued to respond even after the active element segregated away. McClintock argued that these epialleles had predetermined, or ?preset?, patterns of gene activity that were conditioned by the autonomous element 3 , although this response was mostly lost in subsequent generations (erasure). Recent studies on this transposon family have shown that a single active autonomous element can transactivate a second, previously silenced autonomous element, requiring a trans-acting protein that is produced by Spm 126 . Today, presetting is thought to be due to multiple trans- reactivations and modifications of previously silenced autonomous elements that continue to condition the epiallele. REVIEWS 276 | APRIL 2007 | VOLUME 8 www.nature.com/reviews/genetics Constitutive heterochromatin Heterochromatin that is heritable and is found at a particular locus at all times. is involved in silencing P and I TEs (reviewed in REF. 41), but it is not clear whether rasiRNAs regulate all TEs in the germ line, or only those with TE remnants in piRNA clusters. Intriguingly, rasiRNA production and TE silencing are independent of Dicer in D. melanogaster 43 , but at least three helicases ? Homeless/Spindle E, Lighten up and Armitage ? are required, as well as Aubergine and Piwi 41 . Given the unusual single-stranded nature of piRNA, it is likely that these helicases and Argonaute homo- logues cooperate in piRNA biogenesis. Interestingly, similar helicases are found in the chromatin-altering RNA-dependent RNA polymerase (RdRP) silencing complex in fission yeast 52 , suggesting a role for piRNAs in heterochromatin formation. In support of this idea, Piwi and Aubergine are required for some types of PEV (see below) that result in chromatin modifications. TEs contribute to chromosome form and function Centromeres, telomeres and knobs. Constitutive hetero- chromatin is typically found at centromeres and telomeres. Both structures are essential for chromosome function and genome integrity ? the centromere for trafficking chromosomes at cell division, and the telomere for pre- venting chromosome shortening following replication. In most eukaryotic species, the centromere varies in size from several kilobases to several megabases, and consists of long tandem arrays of simple sequence repeats (satel- lite repeats) surrounded by a peripheral region that is rich in TEs. The satellite regions of human centromeres are generally free of TEs, but pericentromeric regions are composed of long blocks of LINE (long interspersed nuclear element) and SINE (short interspersed nuclear element) retrotransposons 53 . In centromeres from A. thaliana, the inner satellite arrays are interspersed with retrotransposons, whereas the outer pericentro- meric region is enriched for DNA transposons 54 . The DNA composition of centromeres differs widely in both length and primary nucleotide sequence between spe- cies, and it is the specific epigenetic context that confers centromeric function (reviewed in REF. 55). The presence of TEs, and the epigenetic silencing apparatus that they recruit, might therefore have a role in centromere func- tion. It is currently unknown whether TEs specifically target heterochromatic regions for transposition, or whether they merely congregate at this location owing to reduced recombination and silencing, although recent data support the second theory 17 . During cell division, the condensed structure of the centromere is required to properly traffic the chromo- some. Reactivation of epigenetically silenced TEs and satellite repeats is accompanied by chromosome segrega- tion and meiotic defects in the mouse 15,56,57 , loss of sister chromatid cohesion in fission yeast (reviewed in REF. 58), and loss of centromere condensation in A. thaliana (reviewed in REF. 59). This suggests that TE silencing mechanisms have a vital role in maintaining centromeric constitutive heterochromatin, and therefore in centro- mere function. In plants, retrotransposons help to silence centromeric satellite arrays (and centromeric satellite arrays help to silence retrotranspons) 60 . Additionally, highly conserved retrotransposons are found exclusively at the centromere, indicating a conserved function (reviewed in REF. 61). In mammals, inactivation of dicer leads to the accumulation of centromeric transcripts 62 , presumably because they are no longer processed into siRNAs. These mutants also have centromeric con- densation and differentiation defects 63 , suggesting that RNAi has an important role in centromere function and heterochromatin formation through the regulation of satellite-repeat and TE transcripts. TEs might have given rise to the centromeric satel- lite tandem repeats themselves. Centromere protein B (CENPB) helps to regulate centromeric heterochroma- tin by binding a short motif in most satellite repeats 64 , and resembles the transposase of Tc1/mariner DNA transposons. Additionally, centromeric satellite repeats from many organisms have sequence homology to known TEs 65 , although they are not mobile and no longer resemble TEs in structure. Tandem repeats such as centromeric satellite repeats are a major source of siRNAs, and it has been suggested that tandem duplica- tions have been retained in the genome to efficiently maintain epigenetic silencing owing to their ability to replenish the siRNA pool 66 . Telomeres are composed of short tandem repeats that, in most species, are added to the ends of chromosomes by the reverse transcriptase telomerase, which uses an RNA template in a mechanism that resembles non-LTR retrotransposon integration 67 . Subtelomeric heterochro- matin that lies adjacent to the telomeric repeats resembles pericentromeric heterochromatin, and is composed of full-length and fragmented TEs. Drosophila species lack a telomerase homologue, and non-LTR retrotransposons have taken over its function by specifically transposing to the chromosome ends in a developmentally regulated and orientation-specific manner (reviewed in REF. 68). These non-LTR retrotransposons (named HeT-A and TART) accomplish this insertion-site specificity by trans- porting their Gag proteins (see BOX 1) into the nucleus, where they recruit more copies to the chromosome ends. Recent studies have shown that these retrotransposons are regulated by the same epigenetic mechanisms that govern TE activity, including RNAi (reviewed in REF. 69). Specifically, mutations in the RNA helicase gene spnE and the argonaute gene aubergine cause an increase in HeT-A and TART transcript levels, reduced siRNA abun- dance and more frequent transpositions to chromosome ends 70 . This RNAi control of HeT-A and TART was found to occur specifically in Drosophila ovaries and oocytes, suggesting that the transposition and subsequent chro- mosome elongation occurs at premeiotic stages of the Drosophila germ line. It is currently unclear whether TEs contribute to telomere-length regulation in mammals. However, recent studies show that the epigenetic status (predominantly DNA methylation) of telomeric and sub-telomeric repeats has a role in regulating telomere elongation in mammalian cells 71 . In contrast to the centromere and telomere, a chromo- somal knob is a fragment of constitutive heterochromatin that has no known function, although some knobs have been associated with neo-centromeric activity. REVIEWS NATURE REVIEWS | GENETICS VOLUME 8 | APRIL 2007 | 277 FOCUS ON EPIGENETICS Analysis of an A. thaliana knob, composed almost entirely of full-length and fragmented TEs (many of which are inserted into each other), showed that hetero- chromatin was dependent on the epigenetic mechanisms of TE silencing, and that the jumbled TE islands that are located in these knobs potentially lead to read-through transcription from existing elements that trigger RNAi- mediated chromatin modifications 16 . This finding, and others, support the view that TE-containing constitutive heterochromatin is governed by the epigenetic regulators that are recruited by TEs. PEV and chromatin insulators. Muller first described PEV in D. melanogaster in the 1930s as a variegated gene-silencing phenotype associated with the spread- ing of heterochromatin into adjacent genes. Transgene insertions in heterochromatin might confer variegated phenotypes (reviewed in REF. 72), in some cases depend- ing on proximity to TEs (FIG. 2a), indicating that TEs are nucleation centres for the formation of this type of facultative heterochromatin 73 . The basis of the variegated phenotype that is seen in PEV is currently enigmatic, although this feature suggests that TE silencing must vary both spatially and temporally. Genetic screens to identify suppressors and enhancers of PEV suggest that the silencing of reporter transgenes is dependent on the same factors that are responsible for silencing TEs, including histone methyltransferases and argo- naute genes, indicating that RNAi-mediated chromatin modifications are involved 74,75 . Genes must be in close proximity to a TE to be influenced by its regulation 16 , and analysis of cis-acting deletions or duplications in D. melanogaster indicate that facultative heterochroma- tin of this sort can spread linearly along the chromosome (for up to 10 kb) until it encounters a boundary before the next actively expressed gene 73 . Chromatin insulators, such as boundary elements, set up territories of gene expression along the chromo- some by establishing domains of chromatin structure. Insulators are thought to have a role in organizing chromatin into nuclear compartments that are depend- ent on transcriptional states (reviewed in REF. 76). TEs themselves can function as insulators. For example, TEs in constitutive heterochromatin might function to distinguish and functionally separate euchromatic chro- mosome arms from telomeric repeats or from the satellite core of the centromere. In D. melanogaster, the Gypsy retrotransposon carries an insulator element that can block a promoter?s interactions with a distal enhancer element, or buffer a transgene from the influence of its genomic context. It remains unresolved whether other TEs besides Gypsy function as insulators. A recent report has shown that, like PEV, Gypsy insulator elements are regulated and controlled by the same set of genes that are responsible for suppression of TEs, including the argonaute family 77 . However, the function of the Gypsy insulator, along with examples of PEV that also require the activity of Piwi and Aubergine, are manifest long after these genes are expressed in the germ line, and must therefore involve chromosomal memory and chromatin modification (reviewed in REF. 41). X inactivation. Female mammals must compensate for their double dosage of X-chromosome genes compared with XY males. This compensation is achieved by the inactivation and heterochromatinization of one of the two X chromosomes. Inactivation is initiated at the X-chromosome inactivation center (XIC), and then spreads outwards from the XIC to the rest of the chro- mosome. However, some regions of the X chromosome are less efficiently silenced. According to one hypothesis, LINE retrotransposons might help as ?boosters? for the efficient spread of the silencing away from the XIC 78 . In support of this idea, the X chromosomes of many mammals, including humans, are rich in LINE elements, except in regions that are prone to escaping X inactiva- tion 79 . Translocations between the X chromosome and autosomes (which have a much lower LINE-element density) show that the spread of silencing away from the XIC and into the autosome cannot be maintained (reviewed in REF. 78). It is still unknown whether LINEs on the X chromosome function in the spread of hetero- chromatin. However, a recent study has shown that short tandem repeats with ancient homology to retrotrans- posons regulate X-chromosome inactivation by producing bidirectional transcripts in differentiating mouse ES cells, providing indirect evidence that TEs function in both the initiation and spread of X inactivation 80 . TE-induced epigenetic regulation of specific genes TEs are known to produce genetic and phenotypic vari- ation between individuals because of their polymorphic locations. They can also produce variation within the cells and tissues of an individual by subjecting nearby genes to the epigenetic regulation that is targeted to TEs. Epialleles are heritable but reversible epigenetic changes in a gene?s expression that are often under the control of a nearby TE or tandem repeat. Epialleles are metastable by nature, having variable expression in genetically identical cells (reviewed in REF. 81), which is manifested as variegation and mosaicism (FIG. 2b) that resembles PEV. As 25% of human promoter regions contain TE-derived sequences, including experimentally verified cis-regulatory elements (reviewed in REF. 82), cryptic TEs might contribute to the regulatory regions of many genes. For example, a recent report showed the transcriptional regulation of the inhibitor of apoptosis protein family in humans, mice and rats were all regulated by domesticated LTR retrotrans- posons. Interestingly, these LTR elements were independ- ently acquired in these species and show no similarity, suggesting that retrotransposons have a regulatory role for orthologous genes and might differentially modulate apoptotic responses between species 83 . Epiallele production and phenotypic variation. TE pro- moters can influence the transcription of nearby genes. In the simplest case, transcripts from full-length TEs can read through the end of the element into a neighbour- ing gene. Read-through transcripts can then be subject to post-transcriptional and transcriptional control, depending on the activity of the TE. Spurious transcrip- tion can also initiate from fractured and incomplete elements that retain promoters but have lost termination Facultative heterochromatin Heterochromatin that switches to euchromatin at points during development. REVIEWS 278 | APRIL 2007 | VOLUME 8 www.nature.com/reviews/genetics MuLE Methylation b Agouti IAP Mu1 Genes TEs 20 kb 0.5 Mb 1.0 Mb hcf106 Dfr-B Variegated expressiona Full expression Figure 2 | The influence of transposable elements on gene expression. a | Position effect variegation (PEV) on Drosophila melanogaster chromosome 4 correlates with the proximity of a transgene to a transposable element (TE) 73 . In this example, expression of the white transgene results in a red eye. Transgene insertions (indicated by triangles) along the chromosome can be near TEs or genes. The phenotype of the eye for each transgene insertion is marked by a red or ?variegated? triangle. Insertions of the transgene into active regions of the genome (such as genes) result in transgene expression and a red eye. Transgene insertions into or near TEs have variable expression resulting in variegated eyes. These insertions demonstrate that variegated transgene expression correlates with proximity to a TE, non-variegated (red-eye) producing transgenes are clustered in gene islands, and the majority of this D. melanogaster chromosome arm is heterochromatic. b | Three examples of variegation induced by TE-generated epialleles. In mice, the intracisternal A-particle (IAP) retrotransposon produces an outward-reading transcript that extends into the agouti coat-colour gene. The level of agouti transcript, and the colour of the coat, is subject to the epigenetic status of the retrotransposon and is heritable 87 . In morning glory flowers, DNA methylation of a non-autonomous MuLE transposon can spread to the promoter of a flower-colour gene (Dfr-B), creating petal-colour streaks 143 . In maize, the activity of one TE family regulates two epialleles. In sectors where the Mutator transposon family is active, the mutant phenotypes of both epialleles occur, generating single sectors of pale green (hcf106 mutant) and necrotic spotted (les28 mutant) tissue on leaves 92 . hcf106 transcripts are initiated by a non-autonomous (Mu1) transposon only when the autonomous transposon elsewhere in the genome is inactive. The molecular nature of the les28 epiallele has not been elucidated. Image in panel a reproduced with permission from REF. 73 ? (2004) American Society for Microbiology. Top image in panel b reproduced with permission from Nature REF. 87 ? (1999) Macmillan Publishers Ltd. Middle image in panel b reproduced with permission from REF. 143 ? (2004) Elsevier Science. Bottom image in panel b reproduced with permission from REF. 144 ? (2001) American Association for the Advancement of Science. REVIEWS NATURE REVIEWS | GENETICS VOLUME 8 | APRIL 2007 | 279 FOCUS ON EPIGENETICS signals. For example, retrotransposon 5? LTRs contain the cis promoter elements that are required to initiate transcription, which are also found in 3? LTRs (BOX 1). Therefore, outward-reading transcripts from retrotrans- posons and solo-LTRs can express downstream ge nes, a process that was first described for Ty1 in budding yeast (reviewed in REF. 84). In addition, DNA transposons, LTR and non-LTR retrotransposons express outward-reading antisense transcripts from their ends. These outward-reading tran- scripts can result in ectopic expression of neighbouring genes that is dependent on the activity state of the TE 85 , as illustrated by the P-cytotype-dependent vestigial locus of D. melanogaster 86 . The mouse IAP LTR retrotransposon produces an outward-reading antisense transcript from its 5? LTR and reads into neighbouring genes, altering their regulation. In one example, a naturally occurring IAP insertion has generated the A vy allele at the mouse agouti locus (FIG. 2b), which confers a different coat colour from the wild-type allele, in addition to increased susceptibility to obesity, diabetes and tumours 87 . In some ways resem- bling presetting (see BOX 3), coat-colour changes that are induced by the IAP element are heritable through the female germ line 88 , and depend on trans-acting modifiers that regulate chromatin and DNA methylation 89 . The human LINE1 element also has an outward- reading promoter in its 5? end, and many human genes use this as their primary promoter 85 . The chimeric transcripts can be tissue specific and potentially expand the expression pattern of the neighbouring gene 90 . Retrotransposons are expressed in mouse cleavage-stage embryos and mature oocytes, providing alternative promoters to a subset of host genes 91 . A recent study showed that the LINE1 outward (antisense) transcript, which overlaps the inward sense transcript, results in dsRNA production and is subject to RNAi, which might bring genes that neighbour the LINE1 element under its epigenetic control 9 . It is also possible that a single TE family can regulate multiple epialleles at the same time 92 . This can result in coordinate epigenetic regulation of several unrelated genes (FIG. 2b), a phe- nomenon that was first described by McClintock. This might suggest that the activity state of a TE, which is subject to regulation by the environment and stress (BOX 4), can potentially regulate a set of genes that have related functions. Furthermore, this highlights the differential regulation of autonomous and non- autonomous elements; although epigenetically silenced, non-autonomous elements are often just responding to the activity of the master autonomous element of the same family. TE-induced epialleles have been proposed to account for phenotypic variability in mammals 93 , especially dif- ferences between genetically identical twins. Variegation and mosaic patterns must begin with a single cell either losing or gaining gene silencing. How the early epi- genetic expression patterns of TEs and epialleles are established differently in identical cells remains a major question in the field. However, because TEs comprise a large fraction of the eukaryotic genome and are the targets of the cell?s gene-silencing machinery, their con- tribution to epigenetic gene regulation and to human disease could be large. Paramutation. Some epialleles undergo paramutation ? an interaction between alleles that results in a her- itable expression change of one allele. The allele with reduced expression (or silenced allele) is then able to initiate silencing of other active alleles in subsequent generations. In maize, several genes in the anthocyanin pigmentation pathway have paramutable alleles, which are associated with and dependent on TEs or tandem repeats. At the R gene, DNA methylation differs between active and silenced alleles, and is localized to a Doppia DNA transposon fragment in the R gene promoter 94 . Intriguingly, in an allele in which most of the transposon fragment had been deleted, cytosine methylation was not present and the ability to change paramutation states was compromised, suggesting that this transposon fragment is responsible for carrying the epigenetic mark. The R homologue B can also be regulated by paramutation, but the regions that are required to control B paramu- tation were mapped to tandem repeats that lie 100 kb upstream from the start of transcription of the B gene 95 . Alleles of B that do not participate in paramutation have only one copy of this repeat, whereas paramutable alleles Box 4 | Genome stress, transposable elements and the adaptive response McClintock first suggested that transposable elements (TEs) were activated in response to challenges to the genome 127 and, since then, the release of the epigenetic silencing of TEs has been described in response to UV exposure, temperature, radiation, wounding, cell culture, pathogen infection and polyploidization (reviewed in REF. 128). Stress-reactivated TEs might generate the raw diversity that a species requires over evolutionary time to survive the specific stress. This adaptive response is a long-term strategy to increase variability for selection, but might not necessarily need to be genetic, as TE-induced epialleles would also be affected if the control of TEs were lost. The response of TEs to stress can occur through one of two mechanisms (reviewed in REF. 128). First, the stress could directly activate TEs and their mutagenic activity. For example, the Tnt1 retrotransposon in tobacco is reactivated by infection, a process that is mediated through the Tnt1 LTR promoter, which has regions that are similar to pathogen defense genes that respond the stress-response hormone salicylic acid. Likewise, some TEs in Drosophila melanogaster that respond to heat stress also contain the same regulatory motifs as heat-shock inducible promoters. Second, stress might inhibit gene-silencing mechanisms in the genome, indirectly resulting in the reactivation of TEs. For example, position effect variegation (PEV) is temperature sensitive in fission yeast 129 as well as in D. melanogaster, as are P element cytotype and hybrid dysgenesis 130 . Additionally, in Drosophila simulans, temperature influences the rate of transposition and, in wild populations, changes in copy number follow a minimum temperature cline 131 . In Schizosaccaromyces pombe, the genes and TEs that are activated in abiotic stress conditions are similar to those that are reactivated in histone deacetylation mutants 23 , and stress-response factors cooperate with RNAi in heterochromatic silencing 132 . These examples demonstrate how the environment can influence the epigenetic regulation of TEs. Once reactivated, these elements can then further alter the genome randomly in response to stress. In one final striking example, dietary supplementation with methyl-group donors for DNA and histone methylation was shown to cause a shift in coat colour in the offspring of A vy mice 133 . Mice that had supplemented diets had a statistically significant shift in the colour of their offspring?s coats compared with non- supplemented control mice. The supplementation altered the level of DNA methylation of the IAP LTR at the A vy locus 134,135 . Genome-wide, it is not known to what extent diet and other aspects of the environment effect epigenetic gene expression that is mediated through TEs, and this remains an important question to be resolved. REVIEWS 280 | APRIL 2007 | VOLUME 8 www.nature.com/reviews/genetics a bcd Immunoglobulin-expressing cells V1 V1 V2 V3 V4 D1 D2 D2 D3 J1 J2 J2 Non-autonomous Transcrib transposonRSS/TIR RAG RAG V2 D1 J2 V4 D2 J2 V1 D3 J1 V2 V3 V4 V1 D2 J2 D1 D3 J1 have seven. DNA methylation and nuclease sensitivity differ at the repeats between active and silenced alleles, and are altered after undergoing paramutation. Recently, an RdRP gene (see BOX 2) was found to be required for paramutation 96 , as well as TEs 97 and transgene silenc- ing 98 . This implicates the RdRP-dependent amplification of non-coding tandem repeat RNAs in the regulation of paramutation. However, paramutation-like phenom- ena also occur in the mouse 99 , which is not thought to have RdRPs. Imprinting. As on the mammalian X chromosome, TEs are specifically associated with monoallelic expressed genes that are located on autosomes, raising the possibil- ity that TEs are involved in the regulation of imprinted genes. In mammals, the IAP and LINE1 retrotransposons are hypomethylated in the female germ line, and consti- tute a high proportion of mRNA in oocytes, which are deposited in the embryo 91 . By contrast, non-autonomous SINE elements are hypomethylated in sperm 100 . This raises the possibility that different classes of TEs might contribute to the differential expression of paternally and maternally imprinted genes. Consistent with this, there is an excess of LINE1 elements that are associ- ated with paternally expressed autosomal imprinted genes 101 , whereas SINE elements are excluded from the promoters of these same genes 102 . X inactivation (see above) in the mouse is also imprinted, such that the paternal X chromosome is preferentially inactivated in the placenta (reviewed in REF. 103), and presumably this could again be due to differential TE activation in the germ line. The effect of TE-mediated imprinting in mammals might be mediated through the action of DNMT3L. This de novo DNA methyltransferase gene is expressed exclusively in female germ line cells and is responsi- ble for DNA methylation and establishment of some genomic imprints 104 . Currently, it is not known how this non-catalytic methyltransferase homologue is guided to TEs, however it might be that stage-specific siRNAs or piRNAs guide DNMT3L to TEs such as IAP and LINE1. Other types of TEs have also given rise to imprinted genes in mammals. For example, the imprinted genes retrotransposon-like 1 (Rtl1) and paternally expressed 10 (Peg10) are immobilized Sushi/Gypsy class retrotrans- posons that have undergone positive selection to acquire new and essential functions 105 . Only the paternal allele of Rtl1 is expressed 106 , and two maternally imprinted microRNAs are encoded on the opposite strand that guide the cleavage of the Rtl1 mRNA 107 . The FWA gene of A. thaliana provides an excel- lent example of how TE-generated tandem repeats are involved in epiallele formation and imprinting. The first two exons of the FWA gene are composed of tandem Figure 3 | The transposable element origins of V(D)J recombination. To generate a diverse set of antibodies, B and T cells of jawed vertebrates undergo V(D)J recombination to create specific DNA rearrangements. a | Immunoglobulin genes are arranged with multiple variable (V), diversity (D) and joining (J) regions, from which genetic diversity is generated upon rearrangement of these regions. Each segment (or exon) is flanked by a short inverted repeat called a recombinational signal sequence (RSS), which is analogous to a DNA transposon?s terminal inverted repeats (TIRs) (BOX 1). The introns that are flanked by RSSs can be thought of as non-autonomous DNA transposons. The two genes that are responsible for the DNA rearrangements, Rag1 and Rag2, are immobilized autonomous versions of the Transcrib family of DNA transposons 145 . These proteins are essentially transposase proteins (BOX 1) that recognize the RSS inverted repeat sequences that are located within the immunoglobulin gene at the site of the coding-sequence boundaries. b | The RAG proteins bind different combinations of RSS sequences in each B or T cell precursor and excise the internal region, which can be thought of as excision of a non-autonomous transposon. c | After excision by the RAG proteins, an immunoglobulin gene is produced with a coding region that has been rearranged compared with the unmodified genome. d | The RSS sequences that interact with the RAG transposase are variable in each cell, giving rise to different excision events and individual immunoglobulin cells that carry different coding potentials. The site of element excision is now a rearranged immunoglobulin gene that can be transcribed and translated into an antibody. REVIEWS NATURE REVIEWS | GENETICS VOLUME 8 | APRIL 2007 | 281 FOCUS ON EPIGENETICS repeats from a fragmented SINE TE. FWA is normally expressed in the seed, and transcription starts in the upstream tandem repeat (which is required for ectopic expression), consistent with a SINE antisense outward- reading promoter. The tandem repeats produce siRNAs that regulate the FWA promoter, and are targeted for DNA methylation 16 . A recent study showed that these tan- dem repeats recruit RNA-dependent DNA methylation to unmethylated FWA epialleles in trans, reminis- cent of paramutation 108 . Normally, FWA is imprinted and expressed from only the maternal allele in the endosperm, in which the tandem repeats (and therefore the SINE) are demethylated 109 . This might reflect the activity of SINE TEs in this tissue; however, TE expres- sion that is dependent on parent of origin has not yet been reported in plants. Mammalian IAP, LINE and SINE retrotransposons show clear parent-of-origin effects (reviewed in REF. 110). However, the extent to which imprinted TEs control imprinted genes remains to be determined. Evidence from plants and animals suggest that the expression patterns of genes that are responsible for epigenetic gene silencing are responsible for generat- ing imprinting patterns. This gene silencing might be mediated through the presence of TEs in or near the imprinted genes. V(D)J Recombination. To generate the enormous diversity of antibodies that are needed for a successful immune response, the B and T cells of jawed vertebrates have co-opted a TE family to create cell-type-specific DNA rearrangements, called V(D)J recombination (FIG. 3). Antibody production and V(D)J recombina- tion are subject to epigenetic regulation, but the extent to which TEs contribute to this regulation is unclear. Because the action of V(D)J recombination must be tissue- and stage-specific, there is tight regulation at the level of recombination signal sequence (RSS) chroma- tin accessibility (reviewed in REF. 111). Non-functional immunoglobulin genes are positioned in the nuclear space close to constitutive heterochromatin, and their position and levels of condensation change upon acti- vation. Histone tail modifications, including methylated H3K9 and chromatin-remodelling by proteins in the SWI2/SNF2 family, regulate the accessibility of the template to V(D)J recombination. Additionally, the process of DNA rearrangement might be under post-transcriptional regulation. Intergenic antisense transcripts have been described that temporally correlate with V(D)J recombination 112 . Therefore, the epigenetic mechanisms that are responsible for regulating V(D)J recombination are reminiscent of those that regulate TE activity. Although it is intriguing that the origin of this rearrangement mechanism is a TE system, whether this origin contributed to the epigenetic regulation of these rearrangements remains unknown. Evolutionary implications TEs have the ability to mutate genes, alter gene regu- lation and generate new genes, each providing fuel for evolution (reviewed in REF. 113). However, this is pre- sumably not the goal of transposition; TEs are likely to be sexually transmitted, self-replicating parasites that do not aim to aid evolution 114,115 . Because the raw variation that is induced by TEs is dependent on their activity, much of the evolutionary potential of TEs is governed by their control and epigenetic regulation. Epigenetic TE regulation might account for some key aspects of genome evolution, including the stress-induced reacti- vation of TEs (BOX 4) and the invasion of a new genome by TEs (BOX 5). TEs also have a major role in generating intraspecies variation. Intergenic regions in different inbred strains of maize, for example, are essentially unrelated owing to differential insertion of LTR-retrotransposons and helitrons 116 . Helitrons are a class of DNA transposon that frequently duplicate passenger genes. In human populations, although active full-length LINE1 ele- ments are undergoing negative selection 117 , they signifi- cantly contribute to levels of human genetic diversity 118 . The Alu family of human SINE elements might also increase intraspecies genetic variability through inser- tions and non-homologous recombination events between elements, leading to various chromosomal rearrangements, duplications and deletions (reviewed in REF. 119). These Alu elements also show epigenetic variability at the level of cytosine methylation of indi- vidual elements 100,120 , producing intraspecies variation. Box 5 | Genome invasion and expansion When a transposable element (TE) enters a genome for the first time (through cross- fertilization, horizontal transfer or another mechanism), the TE can often duplicate freely before becoming epigenetically silenced. For example, engineered versions of Tc1/mariner DNA transposons from frogs and fish 136,137 , or a LINE1 retrotransposon from humans 138 , all had a high transposition frequency when inserted into the mouse genome. Studies in Drosophila melanogaster using the I, hobo and P TEs have provided insight into this phenomenon (reviewed in REF. 139). Older and more fragmented TEs are located primarily in heterochromatin, where they are generally not selected against. It is in this fragmented and epigenetically regulated environment that they are likely to produce the silencing signals (such as small RNAs) that function to silence the entire TE family in trans 41 . As a naive genome does not have these silencing triggers, incoming TE families are able to remain active until they (perhaps by chance) produce an element that functions as a source of the silencing signal. This can occur quickly for TEs that contain inverted repeats, but might also eventually occur due to tandem insertions or abortive transpositions. Using these mechanisms, TEs themselves have been shown to be responsible for their initial epigenetic silencing because of their propensity to generate defective copies upon transposition 123 . TEs also go through bursts of activity during evolution, whenever the restraints on element copy number are released 140 . The nature of these constraints is not known, but epigenetic suppression of TEs probably has a role. Bursts in TE activity can lead to genome size increases, as can genome duplications and polyploidization. For example, large genomes, such as those of humans and maize, consist of ~45% and ~75% TEs, respectively (reviewed in REF. 122). Interestingly, particular families of reactivated retrotransposons provide the bulk of genome size increases in many species. One family of Gypsy-like retrotransposons is responsible for most of the threefold difference in genome size among several Gossypium species 141 . Similarly, as few as three retrotransposon families have doubled the size of some rice genomes 142 . In cases in which only one or a few TEs contribute to genome expansion, the expansion might not be due to total loss of epigenetic control but, instead, loss of control of only one element or family, a process that is similar to what occurs when a TE enters a new genome (see above). For example, the trigger that is responsible for initiating silencing of one TE family might be lost due to deletion or drift, resulting in reactivation of only that family. How particular individual TE families escape silencing or reactivate whereas others do not remains an important question to be resolved. REVIEWS 282 | APRIL 2007 | VOLUME 8 www.nature.com/reviews/genetics Likewise, in plants, meiotically stable epigenetic poly- morphisms between individuals of the same species are often generated by TEs 121 . This intraspecies variation produced by the epigenetic regulation of TEs generates phenotypic variation, and could potentially contribute to speciation itself, through mechanisms such as hybrid vigour or hybrid dysgenesis. Conclusions TEs are a common component in many epigenetic mechanisms, from imprinting and X inactivation to PEV. They have also contributed significantly to the evolution of linear chromosomes, constituting the bulk of pericentromeric and telomeric heterochromatin in most genomes. Likewise, epigenetic regulation of gene expression and chromosome function is strikingly simi- lar to that of TEs, suggesting that they were the original target of this epigenetic regulation. The discovery of transcriptional and post-transcriptional gene-silencing processes that are mediated by RNAi has provided a sequence-specific mechanism that can distinguish TEs from genes. Environmental factors that influence this epigenetic regulation might also have a vital role in disease, natural variation and evolution. Many questions remain. For example, what role do the fractured and jumbled TEs that compose constitutive heterochromatin have in regulating the full-length ele- ments that lie elsewhere in the genome? Additionally, do antiparallel transcripts, such as those that regulate the S. pombe pericentromere, regulate TEs? And how are early expression patterns of TEs and epialleles established differently in identical cells, resulting in variegation? These questions are important and, as we understand more about the epigenetic regulation of TEs, many other epigenetic phenomena are likely to be elucidated because of their common origin. 1. Lander, E. S. et al. Initial sequencing and analysis of the human genome. Nature 409, 860?921 (2001). 2. Girard, L. & Freeling, M. Regulatory changes as a consequence of transposon insertion. Dev. Genet. 25, 291?296 (1999). 3. McClintock, B. Components of action of the regulators Spm and Ac. Carnegie Inst. Wash. Year Book 64, 527?536 (1965). 4. Griffith, J. L. et al. Functional genomics reveals relationships between the retrovirus-like Ty1 element and its host Saccharomyces cerevisiae. Genetics 164, 867?879 (2003). 5. Vastenhouw, N. L. et al. A genome-wide screen identifies 27 genes involved in transposon silencing in C. elegans. Curr. Biol. 13, 1311?1316 (2003). 6. Tabara, H. et al. The rde-1 gene, RNA interference, and transposon silencing in C. elegans. Cell 99, 123?132 (1999). 7. Sijen, T. & Plasterk, R. H. Transposon silencing in the Caenorhabditis elegans germ line by natural RNAi. Nature 426, 310?314 (2003). This report demonstrates that the endogenous RNAi system functions to repress TE activity in a tissue- and developmentally specific manner. 8. Watanabe, T. et al. Identification and characterization of two novel classes of small RNAs in the mouse germline: retrotransposon-derived siRNAs in oocytes and germline small RNAs in testes. Genes Dev. 20, 1732?1743 (2006). This report confirms the previous findings in non-mammalian eukaryotes that TEs are an important source and target of siRNAs. 9. Yang, N. & Kazazian, H. H. Jr. L1 retrotransposition is suppressed by endogenously encoded small interfering RNAs in human cultured cells. Nature Struct. Mol. Biol. 13, 763?771 (2006). This publication demonstrated that siRNAs are generated from TEs in human cells, as well as showing that the LINE1 antisense promoter is essential for the production of these siRNAs. 10. Lippman, Z., May, B., Yordan, C., Singer, T. & Martienssen, R. Distinct mechanisms determine transposon inheritance and methylation via small interfering RNA and histone modification. PLoS Biol. 1, e67 (2003). 11. Gendrel, A. V., Lippman, Z., Yordan, C., Colot, V. & Martienssen, R. A. Dependence of heterochromatic histone H3 methylation patterns on the Arabidopsis gene DDM1. Science 297, 1871?1873 (2002). 12. Martens, J. H. et al. The profile of repeat-associated histone lysine methylation states in the mouse epigenome. EMBO J. 24, 800?812 (2005). 13. Walsh, C. P., Chaillet, J. R. & Bestor, T. H. Transcription of IAP endogenous retroviruses is constrained by cytosine methylation. Nature Genet. 20, 116?117 (1998). 14. Woodcock, D. M., Lawler, C. B., Linsenmeyer, M. E., Doherty, J. P. & Warren, W. D. Asymmetric methylation in the hypermethylated CpG promoter region of the human L1 retrotransposon. J. Biol. Chem. 272, 7810?7816 (1997). 15. Bourc?his, D. & Bestor, T. H. Meiotic catastrophe and retrotransposon reactivation in male germ cells lacking Dnmt3L. Nature 431, 96?99 (2004). This study provides evidence of the tissue-specific de novo methylation of retrotransposons in mammals, as well as demonstrating that this methylation is essential for viability. 16. Lippman, Z. et al. Role of transposable elements in heterochromatin and epigenetic control. Nature 430, 471?476 (2004). This study describes the microarray analysis of an entire chromosomal knob, its TE composition and its dependency on DNA methylation and chromatin remodelling. 17. Kato, M., Takashima, K. & Kakutani, T. Epigenetic control of CACTA transposon mobility in Arabidopsis thaliana. Genetics 168, 961?969 (2004). 18. Yan, Q., Cho, E., Lockett, S. & Muegge, K. Association of Lsh, a regulator of DNA methylation, with pericentromeric heterochromatin is dependent on intact heterochromatin. Mol. Cell. Biol. 23, 8416?8428 (2003). 19. Huang, J. et al. Lsh, an epigenetic guardian of repetitive elements. Nucleic Acids Res. 32, 5019?5028 (2004). This work demonstrates that the SWI/SNF Lsh1 gene regulates only repetitive DNA such as TEs, as well as showing the large amount of the genome that is mis-expressed when TE silencing is lost. 20. Martienssen, R. A., Zaratiegui, M. & Goto, D. B. RNA interference and heterochromatin in the fission yeast Schizosaccharomyces pombe. Trends Genet. 21, 450?456 (2005). 21. Buhler, M., Verdel, A. & Moazed, D. Tethering RITS to a nascent transcript initiates RNAi- and heterochromatin-dependent gene silencing. Cell 125, 873?886 (2006). 22. Irvine, D. V. et al. Argonaute slicing is required for heterochromatic silencing and spreading. Science 313, 1134?1137 (2006). This report demonstrates that the spread of chromatin modifications is dependent on read- through transcription and the catalytic activity of argonaute. 23. Hansen, K. R. et al. Global effects on gene expression in fission yeast by silencing and RNA interference machineries. Mol. Cell. Biol. 25, 590?601 (2005). 24. Mochizuki, K. & Gorovsky, M. A. Small RNAs in genome rearrangement in Tetrahymena. Curr. Opin. Genet. Dev. 14, 181?187 (2004). 25. Xie, Z. et al. Genetic and functional diversification of small RNA pathways in plants. PLoS Biol. 2, e104 (2004). 26. Hamilton, A., Voinnet, O., Chappell, L. & Baulcombe, D. Two classes of short interfering RNA in RNA silencing. EMBO J. 21, 4671?4679 (2002). This is the first correlation between the TE-produced longer class of siRNAs in plants and their role in DNA methylation. 27. Qi, Y. et al. Distinct catalytic and non-catalytic roles of ARGONAUTE4 in RNA-directed DNA methylation. Nature 443, 1008?1012 (2006). 28. Herr, A. J., Jensen, M. B., Dalmay, T. & Baulcombe, D. C. RNA polymerase IV directs silencing of endogenous DNA. Science 308, 118?120 (2005). 29. Onodera, Y. et al. Plant nuclear RNA polymerase IV mediates siRNA and DNA methylation-dependent heterochromatin formation. Cell 120, 613?622 (2005). 30. Huettel, B. et al. Endogenous targets of RNA-directed DNA methylation and Pol IV in Arabidopsis. EMBO J. 25, 2828?2836 (2006). 31. Pontier, D. et al. Reinforcement of silencing at transposons and highly repeated sequences requires the concerted action of two distinct RNA polymerases IV in Arabidopsis. Genes Dev. 19, 2030?2040 (2005). 32. Kanno, T. et al. Atypical RNA polymerase subunits required for RNA-directed DNA methylation. Nature Genet. 37, 761?765 (2005). 33. Li, C. F. et al. An ARGONAUTE4-containing nuclear processing center colocalized with Cajal bodies in Arabidopsis thaliana. Cell 126, 93?106 (2006). 34. Pontes, O. et al. The Arabidopsis chromatin-modifying nuclear siRNA pathway involves a nucleolar RNA processing center. Cell 126, 79?92 (2006). 35. Kanno, T. et al. A SNF2-like protein facilitates dynamic control of DNA methylation. EMBO Rep. 6, 649?655 (2005). 36. Matzke, M. A. & Birchler, J. A. RNAi-mediated pathways in the nucleus. Nature Rev. Genet. 6, 24?35 (2005). 37. Cao, X. & Jacobsen, S. E. Role of the Arabidopsis DRM methyltransferases in de novo DNA methylation and gene silencing. Curr. Biol. 12, 1138?1144 (2002). 38. Weinberg, M. S. et al. The antisense strand of small interfering RNAs directs histone methylation and transcriptional gene silencing in human cells. RNA 12, 256?262 (2006). 39. Castro, J. P. & Carareto, C. M. Drosophila melanogaster P transposable elements: mechanisms of transposition and regulation. Genetica 121, 107?118 (2004). 40. Reiss, D., Josse, T., Anxolabehere, D. & Ronsseray, S. aubergine mutations in Drosophila melanogaster impair P cytotype determination by telomeric P elements inserted in heterochromatin. Mol. Genet. Genomics 272, 336?343 (2004). This report was the first to link an argonaute gene and RNA-based silencing to the trans silencing signal in TE cytotype control. 41. Kavi, H. H., Fernandez, H. R., Xie, W. & Birchler, J. A. RNA silencing in Drosophila. FEBS Lett. 579, 5940?5949 (2005). 42. Jensen, S., Gassama, M. P. & Heidmann, T. Taming of transposable elements by homology-dependent gene silencing. Nature Genet. 21, 209?212 (1999). This important report shows that a short, non- protein-encoding fragment of a TE can suppress hybrid dysgenesis and transposition activity when the full-length TE enters a na�ve genome. REVIEWS NATURE REVIEWS | GENETICS VOLUME 8 | APRIL 2007 | 283 FOCUS ON EPIGENETICS 43. Vagin, V. V. et al. A distinct small RNA pathway silences selfish genetic elements in the germline. Science 313, 320?324 (2006). This recent report shows that rasiRNAs are structurally different from siRNAs, and they associate with Piwi to silence TEs in D. melanogaster. 44. Sarot, E., Payen-Groschene, G., Bucheton, A. & Pelisson, A. Evidence for a Piwi-dependent RNA silencing of the gypsy endogenous retrovirus by the Drosophila melanogaster flamenco gene. Genetics 166, 1313?1321 (2004). 45. Rehwinkel, J. et al. Genome-wide analysis of mRNAs regulated by Drosha and Argonaute proteins in Drosophila melanogaster. Mol. Cell. Biol. 26, 2965?2975 (2006). 46. Cox, D. N. et al. A novel class of evolutionarily conserved genes defined by piwi are essential for stem cell self-renewal. Genes Dev. 12, 3715?3727 (1998). 47. Klattenhoff, C. et al. Drosophila rasiRNA pathway mutations disrupt embryonic axis specification through activation of an ATR/Chk2 DNA damage response. Dev. Cell 12, 45?55 (2007). 48. Aravin, A. et al. A novel class of small RNAs bind to MILI protein in mouse testes. Nature 442, 203?207 (2006). 49. Grivna, S. T., Beyret, E., Wang, Z. & Lin, H. A novel class of small RNAs in mouse spermatogenic cells. Genes Dev. 20, 1709?1714 (2006). 50. Girard, A., Sachidanandam, R., Hannon, G. J. & Carmell, M. A. A germline-specific class of small RNAs binds mammalian Piwi proteins. Nature 442, 199?202 (2006). 51. Aravin, A. A. et al. The small RNA profile during Drosophila melanogaster development. Dev. Cell. 5, 337?350 (2003). 52. Motamedi, M. R. et al. Two RNAi complexes, RITS and RDRC, physically interact and localize to noncoding centromeric RNAs. Cell 119, 789?802 (2004). 53. Schueler, M. G. & Sullivan, B. A. Structural and functional dynamics of human centromeric chromatin. Annu. Rev. Genomics Hum. Genet. 7, 301?313 (2006). 54. The Arabidopsis Genome Initiative. Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408, 796?815 (2000). 55. Dawe, R. K. & Henikoff, S. Centromeres put epigenetics in the driver?s seat. Trends Biochem. Sci. 31, 662?669 (2006). 56. Peters, A. H. et al. Loss of the Suv39h histone methyltransferases impairs mammalian heterochromatin and genome stability. Cell 107, 323?337 (2001). 57. De La Fuente, R. et al. Lsh is required for meiotic chromosome synapsis and retrotransposon silencing in female germ cells. Nature Cell Biol. 8, 1448?1454 (2006). 58. Pidoux, A. L. & Allshire, R. C. The role of heterochromatin in centromere function. Philos. Trans. R. Soc. Lond. B Biol. Sci. 360, 569?579 (2005). 59. Lippman, Z. & Martienssen, R. The role of RNA interference in heterochromatic silencing. Nature 431, 364?370 (2004). 60. May, B. P., Lippman, Z. B., Fang, Y., Spector, D. L. & Martienssen, R. A. Differential regulation of strand- specific transcripts from Arabidopsis centromeric satellite repeats. PLoS Genet. 1, e79 (2005). 61. Jiang, J., Birchler, J. A., Parrott, W. A. & Dawe, R. K. A molecular view of plant centromeres. Trends Plant Sci. 8, 570?575 (2003). 62. Fukagawa, T. et al. Dicer is essential for formation of the heterochromatin structure in vertebrate cells. Nature Cell Biol. 6, 784?791 (2004). 63. Kanellopoulou, C. et al. Dicer-deficient mouse embryonic stem cells are defective in differentiation and centromeric silencing. Genes Dev. 19, 489?501 (2005). This report confirms that, as in lower eukaryotes, dicer and small-RNA-based mechanisms are responsible for centromere condensation in mammals. 64. Ohzeki, J., Nakano, M., Okada, T. & Masumoto, H. CENP-B box is required for de novo centromere chromatin assembly on human alphoid DNA. J. Cell Biol. 159, 765?775 (2002). 65. Wong, L. H. & Choo, K. H. Evolutionary dynamics of transposable elements at the centromere. Trends Genet. 20, 611?616 (2004). 66. Martienssen, R. A. Maintenance of heterochromatin by RNA interference of tandem repeats. Nature Genet. 35, 213?214 (2003). 67. Nakamura, T. M. et al. Telomerase catalytic subunit homologs from fission yeast and human. Science 277, 955?959 (1997). 68. Pardue, M. L. et al. Two retrotransposons maintain telomeres in Drosophila. Chromosome Res. 13, 443?453 (2005). 69. Casacuberta, E. & Pardue, M. L. RNA interference has a role in regulating Drosophila telomeres. Genome Biol. 7, 220 (2006). 70. Savitsky, M., Kwon, D., Georgiev, P., Kalmykova, A. & Gvozdev, V. Telomere elongation is under the control of the RNAi-based mechanism in the Drosophila germline. Genes Dev. 20, 345?354 (2006). 71. Gonzalo, S. et al. DNA methyltransferases control telomere length and telomere recombination in mammalian cells. Nature Cell Biol. 8, 416?424 (2006). 72. Schotta, G., Ebert, A., Dorn, R. & Reuter, G. Position-effect variegation and the genetic dissection of chromatin regulation in Drosophila. Semin. Cell Dev. Biol. 14, 67?75 (2003). 73. Sun, F. L. et al. cis-acting determinants of heterochromatin formation on Drosophila melanogaster chromosome four. Mol. Cell. Biol. 24, 8210?8220 (2004). 74. Haynes, K. A., Caudy, A. A., Collins, L. & Elgin, S. C. Element 1360 and RNAi components contribute to HP1-dependent silencing of a pericentric reporter. Curr. Biol. 16, 2222?2227 (2006). 75. Schotta, G., Ebert, A. & Reuter, G. SU(VAR)3?9 is a conserved key function in heterochromatic gene silencing. Genetica 117, 149?158 (2003). 76. Gaszner, M. & Felsenfeld, G. Insulators: exploiting transcriptional and epigenetic mechanisms. Nature Rev. Genet. 7, 703?713 (2006). 77. Lei, E. P. & Corces, V. G. RNA interference machinery influences the nuclear organization of a chromatin insulator. Nature Genet. 38, 936?941 (2006). 78. Lyon, M. F. Do LINEs have a role in X-chromosome inactivation? J. Biomed. Biotechnol. 2006, 59746 (2006). 79. Ross, M. T. et al. The DNA sequence of the human X chromosome. Nature 434, 325?337 (2005). 80. Cohen, D. E. et al. The DXPas34 repeat regulates random and imprinted X inactivation. Dev. Cell 12, 57?71 (2007). 81. Rakyan, V. K., Blewitt, M. E., Druker, R., Preis, J. I. & Whitelaw, E. Metastable epialleles in mammals. Trends Genet. 18, 348?351 (2002). 82. Jordan, I. K., Rogozin, I. B., Glazko, G. V. & Koonin, E. V. Origin of a substantial fraction of human regulatory sequences from transposable elements. Trends Genet. 19, 68?72 (2003). 83. Romanish, M. T., Lock, W. M., de Lagemaat, L. N., Dunn, C. A. & Mager, D. L. Repeated recruitment of LTR retrotransposons as promoters by the anti- apoptotic locus NAIP during mammalian evolution. PLoS Genet. 3, e10 (2007). 84. Garfinkel, D. J. Genome evolution mediated by Ty elements in Saccharomyces. Cytogenet. Genome Res. 110, 63?69 (2005). 85. Nigumann, P., Redik, K., Matlik, K. & Speek, M. Many human genes are transcribed from the antisense promoter of L1 retrotransposon. Genomics 79, 628?634 (2002). This paper reports the discovery that tissue- specific transcripts that are generated from the outward-reading antisense promoter of LINE1 express neighbouring genes in large numbers. 86. Hodgetts, R. B. & O?Keefe, S. L. The mutant phenotype associated with P-element alleles of the vestigial locus in Drosophila melanogaster may be caused by a readthrough transcript initiated at the P-element promoter. Genetics 157, 1665?1672 (2001). 87. Morgan, H. D., Sutherland, H. G., Martin, D. I. & Whitelaw, E. Epigenetic inheritance at the agouti locus in the mouse. Nature Genet. 23, 314?318 (1999). This landmark report demonstrates that a mosaic coat-colour pattern is defined by the epigenetic state of a nearby TE, and that these epigenetic patterns are not fully erased when passed through meiosis, creating heritable patterns of epigenetic gene expression. 88. Blewitt, M. E., Vickaryous, N. K., Paldi, A., Koseki, H. & Whitelaw, E. Dynamic reprogramming of DNA methylation at an epigenetically sensitive allele in mice. PLoS Genet. 2, e49 (2006). 89. Gaudet, F. et al. Dnmt1 expression in pre- and postimplantation embryogenesis and the maintenance of IAP silencing. Mol. Cell. Biol. 24, 1640?1648 (2004). 90. Matlik, K., Redik, K. & Speek, M. L1 antisense promoter drives tissue-specific transcription of human genes. J. Biomed. Biotechnol. 2006, 71753 (2006). 91. Peaston, A. E. et al. Retrotransposons regulate host genes in mouse oocytes and preimplantation embryos. Dev. Cell 7, 597?606 (2004). 92. Martienssen, R. & Baron, A. Coordinate suppression of mutations caused by Robertson?s mutator transposons in maize. Genetics 136, 1157?1170 (1994). 93. Whitelaw, E. & Martin, D. I. Retrotransposons as epigenetic mediators of phenotypic variation in mammals. Nature Genet. 27, 361?365 (2001). In this provocative commentary, it is suggested that the differential metastable epigenetic silencing of TEs that are scattered throughout the genome produce intraspecies diversity. 94. Walker, E. L. Paramutation of the r1 locus of maize is associated with increased cytosine methylation. Genetics 148, 1973?1981 (1998). 95. Stam, M., Belele, C., Dorweiler, J. E. & Chandler, V. L. Differential chromatin structure within a tandem array 100 kb upstream of the maize b1 locus is associated with paramutation. Genes Dev. 16, 1906?1918 (2002). 96. Alleman, M. et al. An RNA-dependent RNA polymerase is required for paramutation in maize. Nature 442, 295?298 (2006). 97. Woodhouse, M. R., Freeling, M. & Lisch, D. Initiation, establishment, and maintenance of heritable MuDR transposon silencing in maize are mediated by distinct factors. PLoS Biol. 4, e339 (2006). 98. McGinnis, K. M., Springer, C., Lin, Y., Carey, C. C. & Chandler, V. Transcriptionally silenced transgenes in maize are activated by three mutations defective in paramutation. Genetics 173, 1637?1647 (2006). 99. Rassoulzadegan, M. et al. RNA-mediated non- mendelian inheritance of an epigenetic change in the mouse. Nature 441, 469?474 (2006). 100. Rubin, C. M., VandeVoort, C. A., Teplitz, R. L. & Schmid, C. W. Alu repeated DNAs are differentially methylated in primate germ cells. Nucleic Acids Res. 22, 5121?5127 (1994). 101. Allen, E. et al. High concentrations of long interspersed nuclear element sequence distinguish monoallelically expressed genes. Proc. Natl Acad. Sci. USA 100, 9940?9945 (2003). In this unique survey of the genome, LINE elements were found to be significantly associated with imprinted genes. 102. Greally, J. M. Short interspersed transposable elements (SINEs) are excluded from imprinted regions in the human genome. Proc. Natl Acad. Sci. USA 99, 327?332 (2002). 103. Reik, W. & Lewis, A. Co-evolution of X-chromosome inactivation and imprinting in mammals. Nature Rev. Genet. 6, 403?410 (2005). 104. Bourc?his, D., Xu, G. L., Lin, C. S., Bollman, B. & Bestor, T. H. Dnmt3L and the establishment of maternal genomic imprints. Science 294, 2536?2539 (2001). 105. Youngson, N. A., Kocialkowski, S., Peel, N. & Ferguson-Smith, A. C. A small family of sushi-class retrotransposon-derived genes in mammals and their relation to genomic imprinting. J. Mol. Evol. 61, 481?490 (2005). 106. Ono, R. et al. A retrotransposon-derived gene, PEG10, is a novel imprinted gene located on human chromosome 7q21. Genomics 73, 232?237 (2001). 107. Davis, E. et al. RNAi-mediated allelic trans-interaction at the imprinted Rtl1/Peg11 locus. Curr. Biol. 15, 743?749 (2005). 108. Chan, S. W., Zhang, X., Bernatavichute, Y. V. & Jacobsen, S. E. Two-step recruitment of RNA-directed DNA methylation to tandem repeats. PLoS Biol. 4, e363 (2006). 109. Kinoshita, T. et al. One-way control of FWA imprinting in Arabidopsis endosperm by DNA methylation. Science 303, 521?523 (2004). 110. Walter, J., Hutter, B., Khare, T. & Paulsen, M. Repetitive elements in imprinted genes. Cytogenet. Genome Res. 113, 109?115 (2006). 111. Sen, R. & Oltz, E. Genetic and epigenetic regulation of IgH gene assembly. Curr. Opin. Immunol. 18, 237?242 (2006). 112. Bolland, D. J. et al. Antisense intergenic transcription in V(D)J recombination. Nature Immunol. 5, 630?637 (2004). 113. Kidwell, M. G. & Lisch, D. R. Transposable elements and host genome evolution. Trends Ecol. Evol. 15, 95?99 (2000). REVIEWS 284 | APRIL 2007 | VOLUME 8 www.nature.com/reviews/genetics 114. Hickey, D. A. Selfish DNA: a sexually-transmitted nuclear parasite. Genetics 101, 519?531 (1982). 115. Doolittle, W. F. & Sapienza, C. Selfish genes, the phenotype paradigm and genome evolution. Nature 284, 601?603 (1980). 116. Wang, Q. & Dooner, H. K. Eukaryotic transposable elements and genome evolution special feature: remarkable variation in maize genome structure inferred from haplotype diversity at the bz locus. Proc. Natl Acad. Sci. USA 103, 17644?17649 (2006). 117. Boissinot, S., Davis, J., Entezam, A., Petrov, D. & Furano, A. V. Fitness cost of LINE-1 (L1) activity in humans. Proc. Natl Acad. Sci. USA 103, 9590?9594 (2006). 118. Seleme Mdel, C. et al. Extensive individual variation in L1 retrotransposition capability contributes to human genetic diversity. Proc. Natl Acad. Sci. USA 103, 6611?6616 (2006). This report demonstrates that there is a surprisingly high level of individual variation and TE-activity potential between humans. 119. Deininger, P. L., Moran, J. V., Batzer, M. A. & Kazazian, H. H. Jr. Mobile elements and mammalian genome evolution. Curr. Opin. Genet. Dev. 13, 651?658 (2003). 120. Sandovici, I. et al. Interindividual variability and parent of origin DNA methylation differences at specific human Alu elements. Hum. Mol. Genet. 14, 2135?2143 (2005). 121. Rangwala, S. H. et al. Meiotically stable natural epialleles of Sadhu, a novel Arabidopsis retroposon. PLoS Genet. 2, e36 (2006). 122. Kazazian, H. H. Jr. Mobile elements: drivers of genome evolution. Science 303, 1626?1632 (2004). 123. Slotkin, R. K., Freeling, M. & Lisch, D. Heritable transposon silencing initiated by a naturally occurring transposon inverted duplication. Nature Genet. 37, 641?644 (2005). 124. Ronsseray, S., Josse, T., Boivin, A. & Anxolabehere, D. Telomeric transgenes and trans-silencing in Drosophila. Genetica 117, 327?335 (2003). 125. Axtell, M. J., Jan, C., Rajagopalan, R. & Bartel, D. P. A two-hit trigger for siRNA biogenesis in plants. Cell 127, 565?577 (2006). 126. Schlappi, M., Raina, R. & Fedoroff, N. Epigenetic regulation of the maize Spm transposable element: novel activation of a methylated promoter by TnpA. Cell 77, 427?437 (1994). 127. McClintock, B. The significance of responses of the genome to challenge. Science 226, 792?801 (1984). 128. Capy, P., Gasperi, G., Biemont, C. & Bazin, C. Stress and transposable elements: co-evolution or useful parasites? Heredity 85, 101?106 (2000). 129. Allshire, R. C., Javerzat, J. P., Redhead, N. J. & Cranston, G. Position effect variegation at fission yeast centromeres. Cell 76, 157?169 (1994). 130. Kidwell, M. G. Reciprocal differences in female recombination associated with hybrid dysgenesis in Drosophila melanogaster. Genet. Res. 30, 77?88 (1977). 131. Vieira, C., Aubry, P., Lepetit, D. & Biemont, C. A temperature cline in copy number for 412 but not roo/B104 retrotransposons in populations of Drosophila simulans. Proc. Biol. Sci. 265, 1161?1165 (1998). 132. Jia, S., Noma, K. & Grewal, S. I. RNAi-independent heterochromatin nucleation by the stress-activated ATF/CREB family proteins. Science 304, 1971?1976 (2004). 133. Wolff, G. L., Kodell, R. L., Moore, S. R. & Cooney, C. A. Maternal epigenetics and methyl supplements affect agouti gene expression in A vy /a mice. FASEB J. 12, 949?957 (1998). 134. Cooney, C. A., Dave, A. A. & Wolff, G. L. Maternal methyl supplements in mice affect epigenetic variation and DNA methylation of offspring. J. Nutr. 132, 2393S?2400S (2002). This study demonstrates the surprising result that dietary supplementation of a methylation donor molecule heritably alters a TE-induced epiallele that confers coat colour in mice. 135. Waterland, R. A. & Jirtle, R. L. Transposable elements: targets for early nutritional effects on epigenetic gene regulation. Mol. Cell. Biol. 23, 5293?5300 (2003). 136. Ivics, Z., Hackett, P. B., Plasterk, R. H. & Izsvak, Z. Molecular reconstruction of Sleeping Beauty, a Tc1-like transposon from fish, and its transposition in human cells. Cell 91, 501?510 (1997). 137. Miskey, C., Izsvak, Z., Plasterk, R. H. & Ivics, Z. The Frog Prince: a reconstructed transposon from Rana pipiens with high transpositional activity in vertebrate cells. Nucleic Acids Res. 31, 6873?6881 (2003). 138. An, W. et al. Active retrotransposition by a synthetic L1 element in mice. Proc. Natl Acad. Sci. USA 103, 18662?18667 (2006). 139. Biemont, C. & Vieira, C. What transposable elements tell us about genome organization and evolution: the case of Drosophila. Cytogenet. Genome Res. 110, 25?34 (2005). 140. Kidwell, M. G. Transposable elements and the evolution of genome size in eukaryotes. Genetica 115, 49?63 (2002). 141. Hawkins, J. S., Kim, H., Nason, J. D., Wing, R. A. & Wendel, J. F. Differential lineage-specific amplification of transposable elements is responsible for genome size variation in Gossypium. Genome Res. 16, 1252?1261 (2006). 142. Piegu, B. et al. Doubling genome size without polyploidization: dynamics of retrotransposition-driven genomic expansions in Oryza australiensis, a wild relative of rice. Genome Res. 16, 1262?1269 (2006). 143. Iida, S., Morita, Y., Choi, J. D., Park, K. I. & Hoshino, A. Genetics and epigenetics in flower pigmentation associated with transposable elements in morning glories. Adv. Biophys. 38, 141?159 (2004). 144. Martienssen, R. A. & Colot, V. DNA methylation and epigenetic inheritance in plants and filamentous fungi. Science 293, 1070?1074 (2001). 145. Kapitonov, V. V. & Jurka, J. RAG1 core and V(D)J recombination signal sequences were derived from Transib transposons. PLoS Biol. 3, e181 (2005). This study provides the missing evidence of the TE-family origin of the RAG transposase proteins. Acknowledgements The authors apologize if relevant publications were not cited due to space constraints. R.K.S is supported by a postdoc- toral fellowship from the US National Institutes of Health. Work in the authors? laboratory is supported by a grant from the National Institutes of Health to R.M. Competing interests statement The authors declare no competing financial interests. DATABASES The following terms in this article are linked online to: Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query. fcgi?db=gene AGO4 | agouti | Argonaute | CMT3 | DCL3 | Dnmt1 | Dnmt3a | DRD1 | DRM2 | FWA | MET1 | Peg10 | Rag1 | Rag2 | Rtl1 | vestigial UniProtKB: http://ca.expasy.org/sprot Aubergine | CENPB | Clr4 | DDM1 | dicer | DNMT1 | Piwi Access to this links box is available online. REVIEWS NATURE REVIEWS | GENETICS VOLUME 8 | APRIL 2007 | 285 FOCUS ON EPIGENETICS "
Add Content to Group
|
Bookmark
|
Keywords
|
Flag Inappropriate
share
Close
Digg
Facebook
MySpace
Google+
Comments
Close
Please Post Your Comment
*
The Comment you have entered exceeds the maximum length.
Submit
|
Cancel
*
Required
Comments
Please Post Your Comment
No comments yet.
Save Note
Note
View
Public
Private
Friends & Groups
Friends
Groups
Save
|
Cancel
|
Delete
Please provide your notes.
Next
|
Prev
|
Close
|
Edit
|
Delete
Genetics
Gene Inheritance and Transmission
Gene Expression and Regulation
Nucleic Acid Structure and Function
Chromosomes and Cytogenetics
Evolutionary Genetics
Population and Quantitative Genetics
Genomics
Genes and Disease
Genetics and Society
Cell Biology
Cell Origins and Metabolism
Proteins and Gene Expression
Subcellular Compartments
Cell Communication
Cell Cycle and Cell Division
Scientific Communication
Career Planning
Loading ...
Scitable Chat
Register
|
Sign In
Visual Browse
Close
Comments
CloseComments
Please Post Your Comment