Transposable elements as drivers of genomic and biological diversity in vertebrates

Transcription

1 Chromosome Research (2008) 16:203Y215 # Springer 2007 DOI: /s Transposable elements as drivers of genomic and biological diversity in vertebrates Astrid Böhne, Frédéric Brunet, Delphine Galiana-Arnoux, Christina Schultheis & Jean-Nicolas Volff* Institut de Génomique Fonctionnelle de Lyon, Université de Lyon, Institut Fédératif Biosciences Gerland Lyon Sud, Université Lyon 1, CNRS, INRA, Ecole Normale Supérieure de Lyon, France; Tel: +33-(0) ; Fax: +33-(0) ; Jean-Nicolas.Volff@ens-lyon.fr * Correspondence Key words: biodiversity, evolution, genome, transposable element, vertebrate Abstract Comparative genomics has revealed that major vertebrate lineages contain quantitatively and qualitatively different populations of retrotransposable elements and DNA transposons, with important differences also frequently observed between species of the same lineage. This is essentially due to (i) the differential evolution of ancestral families of transposable elements, with evolutionary scenarios ranging from complete extinction to massive invasion; (ii) the lineage-specific introduction of transposable elements by infection and horizontal transfer, as exemplified by endogenous retroviruses; and (iii) the lineage-specific emergence of new transposable elements, as particularly observed for non-coding retroelements called short interspersed elements (SINEs). During vertebrate evolution, transposable elements have repeatedly contributed regulatory and coding sequences to the host, leading to the emergence of new lineage-specific gene regulations and functions. In all vertebrate lineages, there is evidence of transposable element-mediated genomic rearrangements such as insertions, deletions, inversions and duplications potentially associated with or subsequent to speciation events. Taken together, these observations indicate that transposable elements are major drivers of genomic and biological diversity in vertebrates, with possible important roles in speciation and major evolutionary transitions. Abbreviations APE apurinic/apyrimidinic-like endonuclease B-POL type B DNA polymerase CC coiled-coil domain cdna complementary DNA DDE aspartate-aspartate-glutamate EN endonuclease Env envelope L1, LINE1, LINE long interspersed element LTR long terminal repeat MIR mammalian-wide interspersed repeat MITE miniature inverted transposable element MYa million years ago ORF open reading frame PR protease REL REP-HEL RH RPA RT SINE SVA TP VNTR YT Zf restriction enzyme-like endonuclease rolling-circle replication initiator and DNA helicase ribonuclease H ssdna-binding replication protein A-like domain reverse transcriptase short interspersed element FSINE, VNTR and Alu_ target-primed variable number of tandem repeats tyrosine transposase zinc finger

2 204 A. Böhne et al. Introduction Transposable elements are sequences able to Fjump_ into new locations within genomes. They can reach very high copy numbers and represent a major fraction of vertebrate genomes. For instance, over 40% of the human genome is constituted by transposable elements, with a total copy number exceeding that of Fclassical_ genes by about two orders of magnitude (Lander et al. 2001). Transposable elements are a source of mutations and genetic diseases through the induction of various types of rearrangements by transposition and recombination (Deininger et al. 2003, Chen et al. 2005). At first glance, transposable elements behave as Fjunk_ DNA or genomic parasites exploiting host resources for selfish purposes (Doolittle & Sapienza 1980, Orgel & Crick 1980). However, this view has been seriously challenged by a wave of observations supporting a major role of mobile DNA in the structural and functional evolution of genes and genomes in a variety of organisms (Kazazian 2004, Biémont & Vieira 2006). For example, host gene regulatory sequences, polyadenylation signals, coding exons as well as complete RNA genes and open reading frames have been repeatedly derived from transposable elements during evolution, a phenomenon called Fmolecular domestication_, Fexaptation_ or Fco-option_ (van de Lagemaat et al. 2003, Thornburg et al. 2006, Volff 2006, Piriyapongsa et al. 2007, Volff & Brosius 2007). Many of these sequences fulfil biological functions important to the host. For example, the V(D)J recombination-activating protein RAG1, which plays a major role in immune system adaptability in jawed vertebrates, was probably exapted from a transposon about 500 million years ago (Agrawal et al. 1998, Kapitonov & Jurka 2005). Telomerase, a ribonucleoprotein with reverse transcriptase activity catalysing the replication of the ends of eukaryotic chromosomes, might originate from a retrotransposable element (Nakamura & Cech 1998). Therefore, transposable elements serve as a dynamic reservoir of sequences for the evolution of gene function. In addition, mobile DNA has other important general functions, for example in chromosome structure and (epi)genetic regulation (Lyon 2000, Peaston et al. 2004, Han & Boeke 2005, Slotkin & Martienssen 2007). Transposable elements have the potential to shape genomes and to promote genetic innovation, speciation and evolutionary transitions in the vertebrate lineage. Genome drafts and other sequence resources are now available for numerous vertebrate species, allowing one to obtain first insights into the evolutionary impact of transposable elements through comparative genomics. The genomes of several mammals have been already sequenced and published, including human (Lander et al. 2001), chimpanzee Pan troglodytes (Chimpanzee Sequencing And Analysis Consortium 2005), rhesus macaque Macaca mulatta (Rhesus Macaque Genome Sequencing and Analysis Consortium 2007), domestic dog Canis familiaris (Lindblad-Toh et al. 2005), rat Rattus norvegicus (Gibbs et al. 2004), mouse Mus musculus (Mouse Genome Sequencing Consortium 2002) and opossum Monodelphis domestica (marsupial; Mikkelsen et al. 2007). Published genome sequences are also available for the chicken Gallus gallus (International Chicken Genome Sequencing Consortium 2004) and several teleost fishes, including the pufferfish species Takifugu rubripes (Aparicio et al. 2002) and Tetraodon nigroviridis (Jaillon et al. 2004) as well as the medaka Oryzias latipes (Kasahara et al. 2007). Numerous additional projects are very well advanced, with in some cases excellent draft versions already available. This is the case for several mammals (among others the bovine Bos taurus, the horse Equus caballus, the cat Felis catus, the sheep Ovis aries and the pig Sus scrofa), the frog Xenopus tropicalis and several fish species such as the three-spined stickleback Gasterosteus aculeatus and the zebrafish Danio rerio. A significant amount of genomic sequence has been also produced for the elephant shark Callorhinchus milii, a cartilaginous fish (Venkatesh et al. 2007), and genomes of non-vertebrates chordates such as the sea squirt Ciona intestinalis (urochordate; Dehal et al. 2002) are available as Foutgroups_ to orientate evolution. This review aims to integrate data obtained from different species and lineages to reconstruct the evolutionary history of transposable elements in vertebrates and evaluate their potential impact on vertebrate evolution and diversity. Transposable elements in vertebrate genomes The two known major classes of transposable elements, which differ in their mode of transposition, are both represented in vertebrate genomes (Curcio & Derbyshire 2003, Deininger et al. 2003, Kazazian

3 Transposable elements in vertebrates 205 Figure 1. Schematic representation of vertebrate transposable elements. APE, apurinic/apyrimidinic-like endonuclease; B-POL, type B DNA polymerase; CC, coiled-coil domain; EN, endonuclease; ENV, envelope; IN, integrase; LTR, long terminal repeat; MITE, miniature inverted transposable element; ORF, open reading frame; PR, protease; REL, restriction enzyme-like endonuclease; REP-HEL, rolling-circle replication initiator and DNA helicase; RH, ribonuclease H; RPA, ssdna-binding replication protein A-like domain; RT, reverse transcriptase; SINE, short interspersed element; YT, tyrosine transposase; Zf, zinc finger. 2004, Figure 1). Retrotransposable elements (class I), also called retroelements, require for their transposition the reverse transcription of an RNA intermediate into complementary DNA (cdna), while DNA transposons (class II) do not. Retrotransposable elements Autonomous retrotransposable elements encode a reverse transcriptase, an RNA-dependent DNA polymerase catalysing reverse transcription of RNA transcripts into cdna for the production of new copies of the retroelement. Since the original copy is not excised and remains at its location, this process, called retrotransposition, corresponds to a mechanism of duplication. Retrotransposition is largely responsible for the very high copy number reached by retroelements in many vertebrate genomes. Autonomous retroelements are classified according to the presence or absence of long terminal repeats (LTRs). In most LTR retrotransposons and retroviruses, long terminal repeats are arranged in the same orientation and flank the coding part of the element. LTRs are necessary for transcription and cdna integration after reverse transcription. Three major families of LTR retrotransposons have been described in vertebrates: the Ty1/copia, Ty3/gypsy and BEL families (Eickbush & Malik 2002). In these elements, the reverse transcriptase domain is embedded within a larger open reading frame called pol, which encodes a polyprotein with protease (posttranslational processing of retrotransposon proteins), RnaseH (degradation of RNA strand in DNA/RNA heteroduplex) and integrase (integration of the double-stranded cdna after reverse transcription) activities. Integrase is related tovand might be derived from- the DDE transposase of DNA transposons (see below). Another open reading frame called gag is present upstream of pol. Gag frequently overlaps partially with pol and encodes a structural

4 206 A. Böhne et al. protein with nucleic acid binding activity required for assembly of RNA molecules into cytoplasmic retrotransposon particles. Retroviruses additionally encode an envelope (Env) glycoprotein recognizing membrane receptors of the host cell and initiating the process of infection. Vertebrate genomes contain endogenous retroviruses, which have been introduced horizontally during evolution through infection of the germ line. Such elements, generally inactivated by mutations, are then passed vertically to the next generation. Another family of vertebrate LTR retrotransposons, called Y-retrotransposons and related to DIRS1 from the slime mould Dictyostelium discoideum, is more divergent from the structural point of view (Poulter & Goodwin 2005). These elements have more complex repeats and encode, beside an RT/RnaseH protein, a tyrosine transposase instead of the integrase found in other LTR retrotransposons. The Y-transposase has been proposed to catalyse the integration of a circular cdna intermediate produced through reverse transcription (Curcio & Derbyshire 2003). An important group of vertebrate retrotransposons without LTRs is represented by elements alternatively called non-ltr retrotransposons, long interspersed elements (LINEs) or TP (target-primed) retrotransposons (Eickbush & Malik 2002, Curcio & Derbyshire 2003). These elements initiate retrotransposition through a Ftarget-primed_ mechanism: after nicking of the target site by an endonuclease, the released 3 hydroxyl group at the terminal nucleotide of the nick is used to prime reverse transcription (Luan et al. 1993). Vertebrate autonomous non-ltr retrotransposons can encode either an endonuclease related to cellular apurinic/apyrimidinic endonucleases (L1 elements in mammals and other vertebrates), or an endonuclease related to certain restriction enzymes (Rex6 and Zebulon in fish). Another distinct group of autonomous retrotransposons without LTRs is formed by elements related to the Penelope retrotransposon of Drosophila virilis. These elements encode a type of endonuclease also found in mobile introns, and might use target-primed reverse transcription too (Volff et al. 2001a; Pyatkov et al. 2004, Evgen_ev & Arkhipova 2005). Finally, some non-coding retrotransposable elements successfully use the enzymatic machinery encoded by autonomous retroelements to retrotranspose and multiply within vertebrate genomes. Such sequences include short interspersed elements (SINEs) derived from trna, 5S and 7SL RNA genes. This category of non-autonomous retroelements is exemplified by Alu sequences in primates, which use the machinery of L1 non-ltr retrotransposons for their own retrotransposition (Dewannieux et al. 2003). Non-coding LTR retrotransposons can also be mobilized in trans by their autonomous counterparts (Ribet et al. 2004). DNA transposons The second major class of transposable elements in vertebrate genomes is constituted by DNA transposons. Most autonomous DNA transposons encode a so-called DDE transposase, from the triad of highly conserved aspartate-aspartate-glutamate amino-acid residues required for the coordination of metal ions necessary for catalysis (for review see Curcio & Derbyshire 2003). The DDE transposase, which is related to the integrase encoded by some LTR retrotransposons and retroviruses, is responsible for the excision of the element from its original site and its integration into the new target site (Fcut-andpaste_ mechanism). Several superfamilies of Fcutand-paste_ DNA transposons have been identified in vertebrates: Tc1/Mariner, MuDR/Foldback, hat, PiggyBack, PIF, Merlin and P (Feschotte & Pritham 2007). Another type of DNA transposable elements called Helitrons has more recently been identified in vertebrates (Zhou et al. 2006, Kapitonov & Jurka 2007). Helitrons encode a rolling-circle recombinase and presumably paste one strand into the target site and use it as a template for replication. These elements have been shown to transduce host fragments in various organisms (Kapitonov & Jurka 2007). Another group of DNA transposons is constituted by sequences alternatively called Mavericks and Polintons (Kapitonov & Jurka 2006, Pritham et al. 2007). Beside an integrase, these elements encode several proteins similar to proteins from bacteriophages and eukaryotic double-stranded DNA viruses, including a DNA polymerase B homologue, a cystein protease and putative capsid proteins. The transposition mechanism of these elements might involve extrachromosomal replication of a molecule excised from the genome by the polymerase followed by integration of the double-stranded DNA by the integrase. Finally, as observed for retroelements, non-coding DNA transposons can successfully use

5 Transposable elements in vertebrates 207 the enzymatic machinery of autonomous elements for transposition. This is for example the case for MITEs (miniature inverted transposable elements), which are short DNA transposons with terminal inverted repeats using the transposase of autonomous elements (Feschotte & Pritham 2007). Lineage-specific activity and extinction of transposable elements Comparative analyses have demonstrated that genomes in different vertebrate lineages can have very different contents in transposable elements. Certain vertebrate genomes globally contain more transposable elements than others. Mammalian genomes are rich in mobile DNA. However, differences well correlated with variations in haploid genome size are observed between mammals: the genome of the short-tailed opossum contains more transposable elements (52% of the genome) than the genome of primates (45Y50%), mouse and rat (39Y40%) and dog (34%) (Gentles et al and references therein). In birds, the small genome of the domestic chicken (about 40% of the human genome) contains only 10% of mobile DNA (Wicker et al. 2005). In fish, the genome of the medaka (about 25% of the human genome), contains 7% transposable elements (Kasahara et al. 2007). Finally, the pufferfish species Takifugu rubripes and Tetraodon nigroviridis, used as genomic models for the compactness of their genome (12% of the human genome), have only 3Y4% of their DNA constituted by transposable elements (Aparicio et al. 2002, Jaillon et al. 2004). Hence, according to the comparative analysis of sequenced vertebrate genomes, mammalian genomes have the tendency to accumulate more transposable elements. However, interspecific differences within a same lineage are also observed. Variations in copy number are a function of both the activity of the elements and their frequency of elimination. Major differences between vertebrate genomes have been observed concerning the types and families of transposable elements they contain. Such differences can be principally explained by (i) differential transposition success or elimination rate of an ancestral element in different lineages; (ii) horizontal transfer of a transposable element into a specific lineage, for example though infection; or (iii) lineage-specific creation of a new type of transposable element. It should be stressed here that the choice between these different mechanisms is not always obvious. Lineage-specific extinction of transposable elements There is strong evidence for the extinction of ancient families of transposable elements in certain vertebrate lineages. Using a phylogenomic approach, an unexpected diversity of reverse transcriptase-encoding retrotransposons has been observed in fish genomes. More than twenty ancient phylogenetic families of retroelements formed before the split between fish and tetrapods 450 million years ago have been identified in pufferfish and zebrafish (Volff et al. 2003). In contrast, despite a much higher global copy number of retroelements, only three families of retrotransposons are present in mammalian genomes: the non-ltr retrotransposons L1 (LINE1), L2 and L3/CR1. This suggests the elimination of several families of non-ltr retrotransposons with apurinic/apyrimidinic-like endonuclease (for example Rex1, Volff et al. 2000) and restriction-like endonuclease (Rex6 and Zebulon; Volff et al. 2001b, Bouneau et al. 2003), Penelopelike retrotransposons (Volff et al. 2001a, Evgen ev & Arkhipova 2005) and many families of LTR retrotransposons, particularly from the Ty3/gypsy type (Volff et al. 2003). Even for transposable element families present in fish and mammals, fish genomes present a much higher diversity of ancient subfamilies compared with mammals, as reported for L1 (Furano et al. 2004). Many retroelement families identified in fish are also present in amphibian genomes but have been lost in chicken. Taken together, these observations suggest a massive reduction of diversity of retrotransposable elements in the tetrapod lineage, possibly before the split of mammals and birds 300 million years ago. Subsequently, the L1 and L2 families might have been eliminated in the chicken lineage after divergence from mammals (Wicker et al. 2005). Extinction of families of transposable elements can also occur within major vertebrate lineages. For example, the non-ltr retrotransposon Rex3, which is widespread in teleost fish, has been completely lost in salmonids (Volff et al. 2001c). Helitron transposons, which are amplified to over copies in the genome of the bat Myotis lucifugus, are extinct in most other mammals (Pritham & Feschotte 2007).

6 208 A. Böhne et al. Lineage-specific activity of transposable elements The same family of transposable element present in different lineages and species can have very different activity and genomic impact depending on the host. This phenomenon is frequently associated with the differential emergence of variants and subfamilies. The L1 family of non-ltr retrotransposons has literally infested mammalian genomes: copies in the mouse (19% of the genome), in human (17% of the genome) and over 1.1 million copies in the short-tailed opossum (Gentles et al and references therein). In contrast, L1 elements are active but have a much lower copy number in fish (Volff et al. 2003, Furano et al. 2004). Differences in activity can also be observed within a same lineage, even between species with similar number of copies. For example, L1 insertions account for about 2.5% of spontaneous mutations in the mouse, compared with only 0.07% in human. This difference is correlated with the number of intact and potentially retrotransposition-competent elements, estimated at around 3000 in the mouse compared to fewer than 150 in human (Goodier et al. 2001, Mills et al. 2007). However, the L1 family might be on the way to extinction in other rodents (Casavant et al. 2000). Another family with lineagedependent success is L3/CR1. In chicken, this family with copies represents the major group of transposable elements (74% in copy number), covering 3.1% of the genome (Wicker et al. 2005). L3/CR1 elements have been less successful in the larger genome of placental mammals (8000 copies in mouse for 0.05% of the genome and copies in human for 0.31% of the genome) and in fish (Volff et al. 2003), but are well represented in opossum ( copies for 2% of the genome; Gentles et al. 2007). This family of non-ltr retrotransposons was also successful in the genome of the elephant shark Callorhinchus milii (Venkatesh et al. 2007). Finally, Alu SINE elements have been threefold more active in human than in chimpanzee (Chimpanzee Sequencing and Analysis Consortium 2005). Lineage-specific introduction of transposable elements Some retrotransposable elements have been introduced into specific vertebrate lineages through infection and horizontal transfer. This is particularly the case for endogenous retroviruses, which are derived from exogenous retroviruses that have independently infected the germ line of several vertebrate lineages. Evidence for such a recent retroviral invasion of the genome has been observed in the koala (Tarlinton et al. 2006). Endogenous retroviruses can reach very high copy numbers in vertebrate genomes. As much as 5% of the human genome is constituted by about copies of endogenous retroviruses grouped in three classes (Lander et al. 2001), which have been introduced through at least 31 infection events (Belshaw et al. 2005). Most of these elements have subsequently accumulated various types of inactivating mutations, but some of them have conserved a coding potential. Two families of active retroviral elements (PtERV1 and PtERV2) have been introduced into the chimpanzee germ line after divergence from human (Chimpanzee Sequencing and Analysis Consortium 2005). About Y copies of endogenous retroviruses, most of them introduced in a lineagespecific manner, are present in the genome of mouse and opossum, but only copies have been identified in dog and in chicken (Mouse Genome Sequencing Consortium 2002, International Chicken Genome Sequencing Consortium 2004, Lindblad-Toh et al. 2005, Gentles et al. 2007). Endogenous retroviruses are also present in fish, but with modest copy numbers (Jaillon et al. 2004, Shen & Steiner 2004). Furthermore, even non-infectious elements can be transmitted between vertebrate lineages, for example by viral vehicles, as suggested for the transfer of a SINE element from reptiles to mammals using poxviruses (Piskurek & Okada 2007). Horizontal transfer has been also suggested for non-ltr retrotransposons, including Rex1 in fish (Volff et al. 2000) and the RTE family within or between various vertebrate lineages (Kordis & Gubensek 1998, Gentles et al. 2007). Lineage-specific formation of transposable elements Some transposable elements were not present in the last common ancestor of vertebrates and have emerged specifically in some lineages. This is particularly true for non-autonomous transposable elements such as the short interspersed elements (SINEs), which are formed from host RNA genes.

7 Transposable elements in vertebrates 209 Even if some SINE families are ancient (Nishihara et al. 2006), SINEs have generally been formed in a lineage-specific manner. This is the case for the Alu element, a primate-specific SINE derived from 7SL RNA that uses the L1 enzymatic machinery for retrotransposition. Over one million Alu sequences have been identified in human, making up as much as 10% of our genome. Another primate-specific repeat mobilized by L1 is the SVA element. SVA (FSINE, VNTR and Alu_) is a composite non-autonomous retroelement formed by sequences derived from two SINEs, Alu and SINE-R from endogenous retroviral origin, joined by a region containing a variable number of tandem repeats (VNTR region); 2600Y 2700 copies are present in human and chimpanzee genomes. SVA elements emerged before the divergence of all hominids but after the split between hominids and Old World primates (Wang et al. 2005). The mouse genome contains rodent-specific repeats derived from 7SL RNA (B1 repeats) and from trna (B2 repeats), with copies of B1 (2.7% of the genome) and copies of B2 (2.4% of the genome). The dog genome contains a very active family of carnivore-specific SINEs, with about young elements (Lindblad-Toh et al. 2005). SINEs have been also identified in fish, but are rare in chicken. Lineage-specific domestication of transposable element sequences Regulatory and coding sequences from transposable elements (TEs) have repeatedly been recruited during evolution to fulfil host functions. For example, about 4% of human genes have transposable elementderived sequences in their protein-coding regions, and about 25% of human promoter regions contain a TE-derived sequence (Nekrutenko & Li 2001, van de Lagemaat et al. 2003). In addition, numerous RNAand protein-coding genes entirely derived from transposable elements are present in vertebrate genomes. These observations suggest that mobile DNA has significantly contributed to the complexity of vertebrate transcriptomes and proteomes (Horie et al. 2007). Of course, numerous recruited sequences are neutral or deleterious, and will not be fixed by evolution. Particularly, inclusion of a transposable element-derived exon into an open reading frame might introduce frameshift or in-frame premature stop codons and be therefore counterselected. In addition, some domestication events possibly took place before the emergence of the vertebrate lineage, and will therefore not contribute directly to vertebrate diversification. However, there are demonstrated cases of lineage-specific recruitment of transposable element sequences fixed by evolution to perform functions useful to the host. Lineage-specific recruitment of regulatory sequences It has been recently shown that mobile DNA has contributed at least 5.5% of highly conserved mammalian-specific non-exonic sequences, suggesting a major role for transposable elements in shaping and specializing the landscape of gene regulation during mammalian evolution (Lowe et al. 2007). Such potentially exapted regulatory sequences are preferentially located close to genes involved in development and transcription regulation. Several events of lineage-specific recruitment of regulatory sequences have been more particularly studied. This is the case, for example, for the proopiomelanocortin gene POMC, which encodes the prohormone of ACTH (adenocorticotropic hormone), a-, b-, and g-msh (melanocyte-stimulating hormone) and b-endorphin. A neuronal enhancer responsible for the expression of POMC in ventral hypothalamic neurons has been identified in mammals. This enhancer was exapted from an ancient core-sine retroelement more than 170 MYa, before the divergence of placentals, marsupials, and monotremes but was not detected in non-mammalian vertebrates (Santangelo et al. 2007). A functional enhancer derived from a LF-SINE drives the expression of the neurodevelopmental gene ISL1, which encodes a LIM homeobox transcription factor required for motor neuron differentiation. This enhancer is conserved in mammals, chicken and frog, suggesting domestication during the early evolution of the tetrapod lineage (Bejerano et al. 2006). Lineage-specific recruitment of protein-coding exons Protein-coding exons have been also recruited in a lineage-specific manner and conserved in divergent species, indicating a role in the function of the protein. An alternatively spliced exon ultraconserved in placental mammals and opossum, encoding a 31-amino-acid residue sequence, has been recruited by the gene encoding the poly(rc) binding protein 2,

8 210 A. Böhne et al. PCBP2, after divergence of mammalian and bird/ reptile lineages (Bejerano et al. 2006). Another mammalian exon, derived from a MIR (mammalianwide interspersed repeat), is present in the gene for the zinc finger protein 639 (ZNF639), a putative transcriptional repressor. This exon is constitutively expressed in all tested mammals, and no transcript without MIR cassette has been detected (Krull et al. 2007). Exon conservation was observed in placental mammals, marsupials and monotremes but not in non-mammalian vertebrates. Thus, molecular domestication occurred here again after the split between mammalian and bird/reptile lineages. Lineage-specific formation of RNA genes Non-protein-coding RNA genes have been derived from transposable elements in different vertebrate lineages. For example, a neuronal RNA gene called BC200 was formed from an Alu element in a common ancestor of anthropoid primates (including human, chimpanzee, macaque and owl monkey) after divergence from the prosimian lineage (including lemur and galago) (Kuryshev et al. 2001). The function of this gene is unknown, but dysregulation has been observed in certain tumours. Interestingly, the characteristics of BC200 are very reminiscent of those of another neuronal RNA gene called BC1, which has been formed independently in rodents by trna retrotransposition (for review Volff & Brosius 2007). Many microrna (mirna) genes derived from transposable elements have been identified in the human genome, with the potential of regulating thousands of genes (Piriyapongsa et al. 2007). Most of these mirnas are lineagespecific and appeared at different times during the evolution of mammals. Lineage-specific formation of protein-coding genes New lineage-specific protein-coding genes have been derived from transposable elements during vertebrate evolution (Volff 2006). Hundreds of such genes have been identified in the human genome, some of them with important functions in various major biological processes (Zdobnov et al. 2005). These genes have been formed from different open reading frames carried by different types of transposable elements. Beside RAG1, which is present in all jawed vertebrates, several other genes have been derived from DNA transposases, some of them having been fused to zinc finger or other protein domains (Feschotte & Pritham 2007). Several transposase-derived genes are found in most vertebrates, but other are mammal- or even primate-specific. The centromere-associated protein CENP-B, which specifically binds to the centromeric 17 base-pair CENP-B box, is mammalian-specific and was derived from a pogo-like transposase before the divergence of placental mammals, marsupials and monotremes. Interestingly, a similar event of molecular domestication also resulting in the formation of centromeric proteins took place in the fission yeast lineage (Casola et al. 2007). Another mammalspecific pogo-related gene is JERKY, which encodes a brain-specific mrna-binding protein highly expressed in neurons and possibly involved in epilepsy syndromes (Toth et al. 1995). An example of a primate-specific gene is the METNASE/SETMAR chimeric DNA binding protein, which resulted from the fusion of a mariner-type transposase gene with a sequence coding for a SET domain with histone methyltransferase activity. In this case, domestication occurred in a common ancestor of extant anthropoid primates (Cordaux et al. 2006). Various host genes have been derived from retroelements during vertebrate evolution. Particularly, many mammalian genes have been formed from retrotransposon or retroviral gag sequences. As many as 85 gag-like genes grouped in five families have been identified in the human genome. These families have evolved through serial duplications from at least five independent events of molecular domestication (Campillos et al. 2006). The best-studied family of mammal-specific gag-related genes, called MART, is constituted by about 12 members preferentially located on the X chromosome (Brandt et al. 2005). Interestingly, the PEG10 (MART2) gene has been shown to be essential for mouse embryonic development, probably due to a function in placenta formation (Ono et al. 2006). Some MART proteins might work as transcription factors; different functions in the control of cell proliferation and apoptosis, adipocyte differentiation, early embryonic angiogenesis and resistance against retroelement infections have been proposed. Members from other gag-related gene families might be involved in cell survival, death and differentiation (Schuller et al. 2005, Campillos et al. 2006). In mouse, the gag-like gene Fv1 (Friend virus susceptibility 1) restricts murine leukaemia virus replication (Best et al. 1996).

9 Transposable elements in vertebrates 211 Besides gag sequences, envelope (ENV) genes from endogenous retroviruses have also been domesticated in a lineage-dependent manner. Some ENV-like proteins might be involved in the formation of placenta, the nutritional and protective interface between mother and developing fetus in placental mammals. In primates, two envelope-like genes called Syncytin-1 and Syncytin-2 are expressed specifically in placenta and can promote cell fusion as well as syncytium formation in vitro (Mi et al. 2000, Blaise et al. 2003). Accordingly, Syncytin proteins might be involved in the fusion of trophoblast cells leading to the formation of the syncytiotrophoblast layer, a structure with microvillar surfaces facilitating exchanges between mother and fetus. The fusogenic activity of Syncytin-1 is hominoid-specific and conserved in chimpanzee, gorilla, orang-utan and gibbon (Mallet et al. 2004). Syncytin-like genes of retroviral origin with potential roles in placenta formation have also been identified in mouse and sheep (Dupressoir et al. 2005, Dunlap et al. 2006). Interestingly, Syncytin-like genes have been introduced into primate, murine and ovine lineages through independent retroviral infections. Other envelope genes with unknown functions, as well as genes derived from integrases and retroviral-like proteases, are present in mammalian and other vertebrate genomes, where they await further characterization (Krylov & Koonin 2001, Blaise et al. 2003, Volff 2006). Transposable elements and biodiversity in vertebrates Lineage-specific activity and evolution of transposable elements might be a source of biodiversity in vertebrates. Transposable elements are involved in the structure of chromosomes, particularly in centromeres and telomeres. They are also believed to play important roles in different aspects of genetic and epigenetic regulation, including chromatin modification and sex chromosome inactivation (Lyon 2000, Peaston et al. 2004, Han & Boeke 2005, Slotkin & Martienssen 2007). Therefore, lineagespecific expansion of transposable elements might lead to differences in important aspects of genome function and regulation. Differential domestication of regulatory and coding sequences might contribute to the diversity of gene function and regulation, with a possible role in major evolutionary transitions in vertebrates, as suggested by the identification of retroelement-derived genes important for placenta formation (Ono et al. 2006). In some cases, domestication might also be associated with lineagespecific drawbacks. For example, Syncytin-1 mediates the induction of redox reactants, causing oligodendrocyte death and demyelination in human (Antony et al. 2004). Some Gag-derived proteins may be targets of the autoimmune response associated with paraneoplastic neurological disorders in humans (Schuller et al. 2005). Transposition events and transposable elementmediated genomic rearrangements potentially associated with or subsequent to speciation events have been observed in different vertebrate lineages, and are increasingly used as markers to resolve phylogenies (Ray et al. 2006). For example, as many as insertions of transposable elements, principally L1, Alu and SVA, differentiate the human and chimpanzee genomes. Most of them have occurred after divergence from the last common ancestor (Mills et al. 2006) and retrotransposition continues to contribute to human genetic diversity (Seleme et al. 2006). Even DNA transposons, which are not capable of active transposition in the human genome, were intensively active during early primate evolution (Pace & Feschotte 2007). In fish, many bursts of retrotransposition potentially associated with speciation events have been observed for non- LTR retrotransposons (Volff et al. 2000, 2001c). Ectopic recombination between transposable elements can also lead to lineage- or species-specific genomic rearrangements. Some of the nine pericentric inversions distinguishing human and chimpanzee genomes have breakpoints in transposable elements or transposable element-rich regions (for review, Kehrer-Sawatzki & Cooper 2007). Hundreds of deletions took place independently in chimpanzee and human lineages after divergence from their last common ancestor (Sen et al. 2006, Han et al. 2007). Transposable element activity can also lead to the formation of lineage- or species-specific gene duplicates, which can potentially evolve a new regulation or function. This can occur through transposonmediated transduction, as observed for L1 retrotransposons and Helitron transposons (Pickeral et al. 2000, Xing et al. 2006, Pritham & Feschotte 2007). Lineage-specific gene duplicates can also be generated through reverse transcription of cellular genes, resulting in intronless cdna copies called retrogenes. Whereas most of these genes evolved as

10 212 A. Böhne et al. pseudogenes, some copies can acquire a new function (Vinckenbosch et al. 2006). About 200 and 300 retrogenes have been formed in the human and chimpanzee lineages, respectively, after split from the last common ancestor (Chimpanzee Sequencing and Analysis Consortium 2005). Several hypothetical models of speciation might be consistent with a role of transposition and transposable element-mediated rearrangements in speciation. By analogy with hybrid dysgenesis in Drosophila (for review Bucheton 1990), massive transposition in interspecific hybrids might correspond to a mechanism of postzygotic reproductive isolation having negative consequences for hybrid viability or fertility. Alternatively, transposition in hybrids might induce genomic changes leading to speciation. Accordingly, increased transposable expression and transposition have been reported in some plant and animal hybrids (Fontdevila 2005, Noor & Chang 2006 and references therein; O_Neill et al. 1998). Transposition of fertility genes and their subsequent differential genomic localization in two populations has been proposed to cause hybrid sterility and favour speciation (Masly et al. 2006). Finally, a hybrid dysfunction model of chromosomal speciation predicts that differential DNA rearrangement, possibly generated through recombination between transposable elements, will lead to low reproductive fitness in hybrids due to the production of dysfunctional gametes. Alternatively, rearrangements such as inversions might reduce gene flow between populations by suppressing recombination, this eventually leading to speciation (suppressed recombination model; Ayala & Coluzzi 2005). Lineage- and species-specific transposition, gene transduction and DNA rearrangement mediated by vertebrate transposable elements might be consistent with some of these models. Particularly, chromosomal speciation through differential inversions, partly attributable to transposable elements, might have played a role in the divergence between human and chimpanzee (Navarro & Barton 2003). Conclusions Lineage-specific activity, recombination and domestication of transposable elements have probably contributed very significantly to biodiversity in vertebrates, with potential roles in speciation and evolutionary transitions. Several mechanisms might explain lineage-specific differences in transposable content, including differences in host defence, recombination, evolutionary constraints, effective population size and incidence of inbreeding/outcrossing (Neafsey et al. 2004, Abrusan & Krambeck 2006). Some of these processes might also affect not only the quantity and quality of transposable elements, but also their distribution within genomes. For example, an extreme compartmentalization of transposable elements and other repeats is observed in heterochromatic regions of the compact genome of the pufferfish Tetraodon nigroviridis but not in human (Dasilva et al. 2002, Volff et al. 2003). Fitter variants supplanting other transposable elements might also emerge in a lineage-specific manner through adaptive evolution, as observed in vertebrate L1 non-ltr retrotransposons (Boissinot & Furano 2001, Ichiyanagi et al. 2007). Finally, extinction of transposable element families might be linked to molecular domestication events. In many cases, the original transposable element source of the exapted sequence has been eliminated (Volff 2006). This might prevent uncontrolled mobilization in trans and multiplication of the new gene and avoid its integration in unfavourable transcriptional contexts. Even if there is already compelling evidence that transposable elements are major drivers of genomic and biological diversity in vertebrates, our current knowledge probably represents only the tip of the evolutionary iceberg. Most studies performed so far have been done in mammals, and almost nothing is known about the evolutionary impact of transposable elements in fish and other major vertebrate lineages. Even in human and mammals, many aspects of transposable element evolution remain to be studied at the functional level. In the near future, comparative functional genomics will certainly undercover yet unsuspected influences of transposable elements on vertebrate diversity. Acknowledgements Our work is supported by grants from the Ministère de l_education Nationale, de la Recherche et de la Technologie (MENRT), the Fondation pour la Recherche Médicale (FRM), the Association pour la Recherche contre le Cancer (ARC), the Centre National de la Recherche Scientifique (CNRS) and the Institut National de la Recherche Agronomique (INRA).

11 Transposable elements in vertebrates 213 References Abrusan G, Krambeck HJ (2006) Competition may determine the diversity of transposable elements. Theor Popul Biol 70: 364Y375. Agrawal A, Eastman QM, Schatz DG (1998) Transposition mediated by RAG1 and RAG2 and its implications for the evolution of the immune system. Nature 394: 744Y751. Antony JM, van Marle G, Opii W et al. (2004) Human endogenous retrovirus glycoprotein-mediated induction of redox reactants causes oligodendrocyte death and demyelination. Nat Neurosci 7: 1088Y1095. Aparicio S, Chapman J, Stupka E et al. (2002) Whole-genome shotgun assembly and analysis of the genome of Fugu rubripes. Science 297: 1301Y1310. Ayala FJ, Coluzzi M (2005) Chromosome speciation: humans, Drosophila, and mosquitoes. Proc Natl Acad Sci USA 102 (Supplement 1): 6535Y6542. Bejerano G, Lowe CB, Ahituv N et al. (2006) A distal enhancer and an ultraconserved exon are derived from a novel retroposon. Nature 441: 87Y90. Belshaw R, Katzourakis A, Paces J, Burt A, Tristem M (2005) High copy number in human endogenous retrovirus families is associated with copying mechanisms in addition to reinfection. Mol Biol Evol 22: 814Y817. Best S, Le Tissier P, Towers G, Stoye JP (1996) Positional cloning of the mouse retrovirus restriction gene Fv1. Nature 382: 826Y829. Biémont C, Vieira C (2006) Genetics: junk DNA as an evolutionary force. Nature 443: 521Y524. Blaise S, de Parseval N, Benit L, Heidmann T (2003) Genomewide screening for fusogenic human endogenous retrovirus envelopes identifies syncytin 2, a gene conserved on primate evolution. Proc Natl Acad Sci USA 100: 13013Y Boissinot S, Furano AV (2001) Adaptive evolution in LINE-1 retrotransposons. Mol Biol Evol 18: 2186Y2194. Bouneau L, Fischer C, Ozouf-Costaz C et al. (2003) An active non-ltr retrotransposon with tandem structure in the compact genome of the pufferfish Tetraodon nigroviridis. Genome Res 13: 1686Y1695. Brandt J, Schrauth S, Veith AM et al. (2005) Transposable elements as a source of genetic innovation: expression and evolution of a family of retrotransposon-derived neogenes in mammals. Gene 345: 101Y111. Bucheton A (1990) I transposable elements and I-R hybrid dysgenesis in Drosophila. Trends Genet 6: 16Y21. Campillos M, Doerks T, Shah PK, Bork P (2006) Computational characterization of multiple Gag-like human proteins. Trends Genet 22: 585Y589. Casavant NC, Scott L, Cantrell MA, Wiggins LE, Baker RJ, Wichman HA (2000) The end of the LINE?: lack of recent L1 activity in a group of South American rodents. Genetics 154: 1809Y1817. Casola C, Hucks D, Feschotte C (2007) Convergent domestication of pogo-like transposases into centromere-binding proteins in fission yeast and mammals. Mol Biol Evol (Oct 16); [Epub ahead of print]. Chen JM, Stenson PD, Cooper DN, Ferec C (2005) A systematic analysis of LINE-1 endonuclease-dependent retrotranspositional events causing human genetic disease. Hum Genet 117: 411Y427. Chimpanzee Sequencing and Analysis Consortium (2005) Initial sequence of the chimpanzee genome and comparison with the human genome. Nature 437: 69Y87. Cordaux R, Udit S, Batzer MA, Feschotte C (2006) Birth of a chimeric primate gene by capture of the transposase gene from a mobile element. Proc Natl Acad Sci USA 103: 8101Y8106. Curcio MJ, Derbyshire KM (2003) The outs and ins of transposition: from mu to kangaroo. Nat Rev Mol Cell Biol 4: 865Y877. Dasilva C, Hadji H, Ozouf-Costaz C et al. (2002) Remarkable compartmentalization of transposable elements and pseudogenes in the heterochromatin of the Tetraodon nigroviridis genome. Proc Natl Acad Sci USA 99: 13636Y Dehal P, Satou Y, Campbell RK et al. (2002) The draft genome of Ciona intestinalis: insights into chordate and vertebrate origins. Science 298: 2157Y2167. Deininger PL, Moran JV, Batzer MA, Kazazian HH Jr (2003) Mobile elements and mammalian genome evolution. Curr Opin Genet Dev 13: 651Y658. Dewannieux M, Esnault C, Heidmann T (2003) LINE-mediated retrotransposition of marked Alu sequences. Nat Genet 35: 41Y48. Doolittle WF, Sapienza C (1980) Selfish genes, the phenotype paradigm and genome evolution. Nature 284: 601Y603. Dunlap KA, Palmarini M, Varela M et al. (2006). Endogenous retroviruses regulate periimplantation placental growth and differentiation. Proc Natl Acad Sci USA 103: 14390Y Dupressoir A, Marceau G, Vernochet C et al. (2005) Syncytin-A and syncytin-b, two fusogenic placenta-specific murine envelope genes of retroviral origin conserved in Muridae. Proc Natl Acad Sci USA 102: 725Y730. Eickbush TH, Malik HS (2002) Origins and evolution of retrotransposons. In: Craig NL, Craigie R, Gellert M, Lambowitz AM, eds., Mobile DNA II, ASM Press, Washington, pp. 1111Y1144. Evgen_ev MB, Arkhipova IR (2005) Penelope-like elements Y a new class of retroelements: distribution, function and possible evolutionary significance. Cytogenet Genome Res 110: 510Y521. Feschotte C, Pritham EJ (2007) DNA transposons and the evolution of eukaryotic genomes. Annu Rev Genet 41: 331Y368. Fontdevila A (2005) Hybrid genome evolution by transposition. Cytogenet Genome Res 110: 49Y55. Furano AV, Duvernell DD, Boissinot S (2004) L1 (LINE-1) retrotransposon diversity differs dramatically between mammals and fish. Trends Genet 20: 9Y14. Gentles AJ, Wakefield MJ, Kohany O et al. (2007) Evolutionary dynamics of transposable elements in the short-tailed opossum Monodelphis domestica. Genome Res 17: 992Y1004. Gibbs RA, Weinstock GM, Metzker ML et al. (2004) Genome sequence of the Brown Norway rat yields insights into mammalian evolution. Nature 428: 493Y521. Goodier JL, Ostertag EM, Du K, Kazazian HH Jr (2001) A novel active L1 retrotransposon subfamily in the mouse. Genome Res 11: 1677Y1685. Han JS, Boeke JD (2005) LINE-1 retrotransposons: modulators of quantity and quality of mammalian gene expression? Bioessays 27: 775Y784. Han K, Lee J, Meyer TJ et al. (2007) Alu recombination-mediated structural deletions in the chimpanzee genome. PLoS Genet 3: e184.

12 214 A. Böhne et al. Horie K, Saito ES, Keng VW, Ikeda R, Ishihara H, Takeda J (2007) Retrotransposons influence the mouse transcriptome: implication for the divergence of genetic traits. Genetics 176: 815Y827. Ichiyanagi K, Nishihara H, Duvernell DD, Okada N (2007) Acquisition of endonuclease specificity during evolution of L1 retrotransposon. Mol Biol Evol 24: 2009Y2015. International Chicken Genome Sequencing Consortium (2004) Sequence and comparative analysis of the chicken genome provide unique perspectives on vertebrate evolution. Nature 432: 695Y716. Jaillon O, Aury JM, Brunet F et al. (2004) Genome duplication in the teleost fish Tetraodon nigroviridis reveals the early vertebrate proto-karyotype. Nature 431: 946Y957. Kapitonov VV, Jurka J (2005) RAG1 core and V(D)J recombination signal sequences were derived from Transib transposons. PLoS Biol 3: e181. Kapitonov VV, Jurka J (2006) Self-synthesizing DNA transposons in eukaryotes. Proc Natl Acad Sci USA 103: 4540Y4545. Kapitonov VV, Jurka J (2007) Helitrons on a roll: eukaryotic rolling-circle transposons. Trends Genet 23: 521Y529. Kasahara M, Naruse K, Sasaki S et al. (2007) The medaka draft genome and insights into vertebrate genome evolution. Nature 447: 714Y719. Kazazian HH Jr (2004) Mobile elements: drivers of genome evolution. Science 303: 1626Y1632. Kehrer-Sawatzki H, Cooper DN (2007) Structural divergence between the human and chimpanzee genomes. Hum Genet 120: 759Y778. Kordis D, Gubensek F (1998) Unusual horizontal transfer of a long interspersed nuclear element between distant vertebrate classes. Proc Natl Acad Sci USA 95: 10704Y Krull M, Petrusma M, Makalowski W, Brosius J, Schmitz J (2007) Functional persistence of exonized mammalian-wide interspersed repeat elements (MIRs). Genome Res 17: 1139Y1145. Krylov DM, Koonin EV (2001) A novel family of predicted retroviral-like aspartyl proteases with a possible key role in eukaryotic cell cycle control. Curr Biol 11: R584YR587. Kuryshev VY, Skryabin BV, Kremerskothen J, Jurka J, Brosius J (2001) Birth of a gene: locus of neuronal BC200 snmrna in three prosimians and human BC200 pseudogenes as archives of change in the Anthropoidea lineage. J Mol Biol 309: 1049Y1066. Lander ES, Linton LM, Birren B et al. (2001) Initial sequencing and analysis of the human genome. Nature 409: 860Y921. Lindblad-Toh K, Wade CM, Mikkelsen TS et al. (2005) Genome sequence, comparative analysis and haplotype structure of the domestic dog. Nature 438: 803Y819. Lowe CB, Bejerano G, Haussler D (2007) Thousands of human mobile element fragments undergo strong purifying selection near developmental genes. Proc Natl Acad Sci USA 104: 8005Y8010. Luan DD, Korman MH, Jakubczak JL, Eickbush TH (1993) Reverse transcription of R2Bm RNA is primed by a nick at the chromosomal target site: a mechanism for non-ltr retrotransposition. Cell 72: 595Y605. Lyon MF (2000) LINE-1 elements and X chromosome inactivation: a function for Bjunk^ DNA? Proc Natl Acad Sci USA 97: 6248Y6249. Mallet F, Bouton O, Prudhomme S et al. (2004) The endogenous retroviral locus ERVWE1 is a bona fide gene involved in hominoid placental physiology. Proc Natl Acad Sci USA 101: 1731Y1736. Masly JP, Jones CD, Noor MA, Locke J, Orr HA (2006) Gene transposition as a cause of hybrid sterility in Drosophila. Science 313: 1448Y1450. Mi S, Lee X, Li XP et al. (2000). Syncytin is a captive retroviral envelope protein involved in human placental morphogenesis. Nature 403: 785Y789. Mikkelsen TS, Wakefield MJ, Aken B et al. (2007) Genome of the marsupial Monodelphis domestica reveals innovation in noncoding sequences. Nature 447: 167Y177. Mills RE, Bennett EA, Iskow RC et al. (2006) Recently mobilized transposons in the human and chimpanzee genomes. Am J Hum Genet 78: 671Y679. Mills RE, Bennett EA, Iskow RC, Devine SE (2007) Which transposable elements are active in the human genome? Trends Genet 23: 183Y191. Mouse Genome Sequencing Consortium (2002) Initial sequencing and comparative analysis of the mouse genome. Nature 420: 520Y562. Nakamura TM, Cech TR (1998) Reversing time: origin of telomerase. Cell 92: 587Y590. Navarro A, Barton NH (2003) Chromosomal speciation and molecular divergence - accelerated evolution in rearranged chromosomes. Science 300: 321Y324. Neafsey DE, Blumenstiel JP, Hartl DL (2004) Different regulatory mechanisms underlie similar transposable element profiles in pufferfish and fruitflies. Mol Biol Evol 21: 2310Y2318. Nekrutenko A, Li WH (2001) Transposable elements are found in a large number of human protein-coding genes. Trends Genet 17: 619Y621. Nishihara H, Smit AF, Okada N (2006) Functional noncoding sequences derived from SINEs in the mammalian genome. Genome Res 16: 864Y874. Noor MA, Chang AS (2006) Evolutionary genetics: jumping into a new species. Curr Biol 16: R890YR892. O_Neill RJ, O_Neill MJ, Graves JA (1998) Undermethylation associated with retroelement activation and chromosome remodelling in an interspecific mammalian hybrid. Nature 393: 68Y72. Ono R, Nakamura K, Inoue K et al. (2006) Deletion of Peg10, an imprinted gene acquired from a retrotransposon, causes early embryonic lethality. Nat Genet 38: 101Y106. Orgel LE, Crick FH (1980) Selfish DNA: the ultimate parasite. Nature 284: 604Y607. Pace JK II, Feschotte C (2007) The evolutionary history of human DNA transposons: evidence for intense activity in the primate lineage. Genome Res 17: 422Y432. Peaston AE, Evsikov AV, Graber JH et al. (2004). Retrotransposons regulate host genes in mouse oocytes and preimplantation embryos. Dev Cell 7: 597Y606. Pickeral OK, Makalowski W, Boguski MS, Boeke JD (2000) Frequent human genomic DNA transduction driven by LINE-1 retrotransposition. Genome Res 10: 411Y415. Piriyapongsa J, Marino-Ramirez L, Jordan IK (2007) Origin and evolution of human micrornas from transposable elements. Genetics 176: 1323Y1337.