Detection of in vivo Transposition Events by TE Display
TE display (Van den Broeck et al., 1998; Casa et al., 2000; Biedler et al., 2003) is a sensitive and reproducible experimental method to detect TE insertions (Figure 5). TE display is a powerful tool for genome-wide analysis of TE insertions, and for detection of new insertions due to transposition (De Keukeleire et al., 2001). It offers a higher degree of sensitivity and resolution than genomic Southern analysis. TE display has been used to detect somatic transposition (De Keukeleire et al., 2001; Sethuraman et al. , 2007). This is done simply by looking for the presence of new bands that represent newly transposed copies of a TE, although a caveat of this approach is that change of restriction site may also result in new TE display bands. The same method may also be used to identify germ-line transposition by comparing TE display patterns of parent insects with the patterns of a large number of offspring. Alternatively, one could take advantage of the possibility that some TEs are activated in cell culture. Using TE display, it may be possible to identify active families by comparing the relative abundance of a TE in cultured cells with that in individuals from different strains of the same species (Jiang et al., 2003). This approach is based on the assumption that some TE families may be more active in cultured cells than in live organisms.
Detection of in vivo Transposition Events by Inverse PCR
Actively transposing DNA-mediated TEs can be identified as extrachromsomal DNA in the form of linear or circular intermediates or byproducts (Arca et al., 1997; Gorbunova and Levy, 1997; O’Brochta et al, 2009). Using a set of outward-orienting primers within the TE, the circular extrachromosomal copies may be amplified, which may serve as evidence of active excision or transposition. However, head-to-head copies of the same TE in the genome could also produce PCR products when outward-orienting primers are used, which need to be ruled out by sequencing and further analysis.
Transposition Assay, Reconstruction, and Genetic Screen
Transposition assays can be used to directly assess the functionalities of both the cis- (TIRs) and the trans-(transposase) components of a DNA transposon, allowing the demonstration of autonomous transposition events.In addition, transposition assays have also been established for the detection of retrotransposition of non-LTR retrotransposons (Jensen et al., 1994; Ostertag et al., 2000). A molecular reconstruction approach has been developed to restore inactivated copies of vertebrate transposons named Sleeping Beauty (Ivics et al., 1997) and Harbinger (Sinzelle et al., 2008). This approach may become increasingly feasible as the cost of gene synthesis continues to drop. As an alternative, a genetic screen based on a bacterial system has been developed to identify hyperactive copies of an insect mariner transposon among randomly mutated copies (Lampe et al., 1999). This approach can potentially be used to screen for active copies of transposons that do not require specific host factors. In summary, the progress in insect genome projects and the development and application of the methods described in this section will greatly facilitate searches for active TEs. The task of finding needles in the haystack could potentially be replaced by targeted and efficient investigations.
Evolution of Insect TEs
The evolutionary dynamics of TEs are complex, in part because of their replication and their interactions with the host genome. The intricate dynamics between TEs and their host genomes are further complicated by the ability of some TEs to cross species barriers and spread to the genome of a new species by horizontal transfer. Horizontal transfer may be an important part of the life cycle of some TEs, and contribute to their continued success during evolution (Silva et al., 2004; Schaack et al., 2010). While the broad distribution of both RNA-mediated TEs and DNA-mediated TEs in all eukaryotic groups is evidence of the long-term evolutionary success of TEs, different TEs may have adopted different strategies, for which several insect TEs in both classes provide good examples.
Genomic Considerations of TE Evolution
It has been hypothesized that TE insertions may present three types of potential deleterious effects, including: (1) insertional mutagenesis, which may disrupt gene function and/or regulation; (2) transcriptional/translational cost of the production of TE transcripts and proteins; and (3) ectopic recombination between homologous copies of TEs in different chromosomal locations, which may result in duplication, deletion, and new linkage relationships between genes (Nuzhdin, 1999; Kidwell and Lisch, 2001; Bartolome et al., 2002; Petrov et al., 2003; Feschotte and Pritham, 2007). The costs of having TEs may also include the costs associated with DNA replication when TEs occupy a large fraction of the genome. Obviously, these hypotheses are not mutually exclusive. This section discusses the intra-genomic dynamics of TE—host interaction. The population dynamics affecting the spread of TEs in insects, which is also important for TE evolution, will be discussed in section 3.7.
Self-regulation of insect TEs Self-regulation has been shown for Drosophila mariner and P elements (Hartl et al, 1997; Kidwell and Lisch, 2001). In the case of Drosophila P element, self-regulation is achieved through the activities of at least two types of element-encoded repressors. In the case of mariner, several mechanisms may be involved, including overproduction inhibition (an increase in the amount of transposase results in a decrease in net transposase activity), missense mutation effects (defective transposase encoded by missense copies interfering with functional transposase), and titration effects by inactive copies. In this regard, it is interesting to note that several hyperactive mutants of an active mariner, originally discovered from the horn fly, have been isolated (Lampe et al., 1999). This suggests that the horn fly mariner has not evolved for maximal activity.
Host-control of insect TEs and small RNA pathways RNA interference (RNAi), a mechanism that confers post-transcriptional suppression on the basis of homology to small fragments of double-stranded RNA, has long been implicated as a host defense mechanism against a broad spectrum of TEs in the nemotode Caenorhabditis elegans (Ketting et al., 1999; Tabara et al., 1999). Recent studies in Drosphila uncovered diverse and complex small RNA pathways that control TEs in germ-line and somatic tissues (reviewed in Malone and Hannon, 2009; Forstemann, 2010). The invovement of small RNAs in the control of TE activity in Drosophila was initially proposed on the basis of cosuppression of the I element by an increasing number of I-related transgenes (Jensen et al., 1999; Labrador and Corces, 2002). Repeat-associated small interfering RNAs (rasiRNA) were discovered in the Drosphila germ-line; these RNAs bind Piwi-type proteins (Piwi and Aubergine), and were later called Piwi-interacting RNAs, or piRNAs (Vagin et al., 2006; Brennecke et al., 2007; Gunawardane et al., 2007). These piRNAs correspond to TEs in different groups, including roo, I, and gypsy. Piwi and aubergine mutant flies were shown to de-repress gypsy, TART, and P elements (reviewed in Malone and Hannon, 2009). piRNAs mostly originate from piRNA clusters, which consist of many truncated, nested, and inactivated TE remnants in the Drosophila genome. In the germ-line, piRNAs are amplified through a ping-pong mechanism involving Piwi, Aubergine, and Ago3. piRNAs are also found in the somatic support cells of the ovary, where no ping-pong amplification is detected because of the lack of Aubergine and Ago3 (Malone et al., 2009; Li et al., 2009). In the somatic support cells, these piRNAs correspond to sequences that reside in the flamenco/COM locus, which suppress gypsy, ZAM, and idefix elements (Pelisson et al., 2007; Desset et al., 2008). These observations link one of the best early examples of host TE control (flamenco over gypsy, Bucheton, 1995) with the small RNA pathway. piRNAs may be inherited maternally, conferring TE suppression to the offspring epigenetically. This epigenetic control was shown in the P- and I-element-mediated hybrid dysgenesis (Brennecke et al., 2008). In essence, when males that contain active P or I elements mate with females that have no piRNAs against these TEs, reduced fertility will result as a consequence of uncontrolled P- or I-element transposition in the offspring.
piRNAs differ from the originally discovered siRNAs because piRNAs are longer in their length (24-29 nt versus 21-22 nt) and bind Piwi-type proteins. Endogenous siRNAs (endo-siRNAs) that correspond to TEs and other repetitive sequences have also been found in the Dro-sophila germ-line and somatic cells (Czech et al., 2008; Ghildiyal et al., 2008; Kawamura et al., 2008; Okamura et al., 2008). These endo-siRNAs may be derived from sense and antisense transcription of repeats as well as host genes, which may regulate gene expression and repress TE activity, depending on the specific target. There is also an indication that an RNA-dependent RNA polymerase in Drosophila may produce dsRNAs that will be processed to make endo-siRNAs (Lipardi and Paterson, 2009).
Two other points related to host control of TE activities are worth noting. The first is the link between TE silencing by piRNAs and heterochromatin. Klattenhoff and colleagues (2009) recently reported that an HP1 (hetero-chromatin protein 1) family protein is required for transposon silencing, piRNA production and amplification. The second significant discovery is that the alterations of the Hsp90 chaperone machinery affect the piRNA pathway and lead to transposon activation and mutation (Specchia et al., 2010). The authors hypothesize that Hsp90 may act as a genetic buffering system for TE activity, thus supporting the concept of canalization during development.
Non-random distribution of insect TEs Patterns of non-random TE distribution have been shown in both D. melanogaster and An. gambiae (Bartolome et al., 2002; Holt et al., 2002; Kapitonov and Jurka, 2003a). TEs tend to accumulate in heterochromatin. Such a distribution bias could result either from preferential TE insertion, or from selection against insertions in euchromatic regions, or from both. Bartolome and colleagues suggest that the abundance of TEs is more strongly associated with local recombination rates (Bartolome et al., 2002), which are low in heterochromatic regions, rather than with gene density. They argue that this association is consistent with the hypothesis that selection against harmful effects of ectopic recombination is a major force opposing TE spread. However, selection against insertional mutagenesis is also at work, as shown by the absence of insertions in coding regions. The insertional bias of P elements has been recently demonstrated during genome-scale P mutagenesis analysis (Spradling et al., 1995, 1999). Therefore, insertion bias may contribute to the biased pattern of TE distribution in insects. A related topic here is the suggestion that concentrations of TE insertions in the Drosophila Y-chromosome may have contributed to the evolutionary process leading to its inactivation (Labrador and Corces, 2002). It should be noted that not all TEs have a bias towards heterochromatic or recombination-deprived regions. On the basis of analysis of limited gene sequences, it was shown that Ae. aegypti MITEs tend to be associated with the non-coding regions within or near genes (Tu, 1997), which is similar to what has been observed for plant MITEs (Zhang et al., 2000).
Autonomous and non-autonomous TEs In addition to the genomic interactions described above, most TEs have to contend with the fact that defective copies are often generated during or after transposition. This process could contribute to self-regulation, as discussed above. It can also lead to total inactivation and ultimate extinction of a TE, as the inactive TE population eventually overwhelms the active copies (Eickbush and Malik, 2002; Feschotte and Pritham, 2007). Therefore, the replicative ability that is responsible for the success of the TE may also lead to its inactivation in a genome. Interestingly, some non-autonomous TEs have been very successful with regard to amplification, although the mechanisms that contribute to their success are not entirely clear. For example, SINEs found in insect genomes, including Ae. aegypti and B. mori, all contain thousands of copies (Adams et al., 1986; Tu, 1999).
Figure 6 A model of the evolutionary dynamics of TEs in eukaryotic genomes. This hypothetical model incorporates recent work by several groups (Hartl et al., 1997; Lampe et al., 2001; Silva et al., 2004). Some aspects of the model are better suited for DNA transposons that generally have a high propensity for horizontal transfer.
Similarly, most of the MITEs found in insects are also highly reiterated, although low copy-number families are also found (Tu, 1997, 2000, 2001a; Nene et al., 2007; Han et al., 2010). The small size of MITEs and SINEs may confer less deleterious effects on the host, either because they are less efficient substrates for homologous recombination (Petrov et al., 2003) or because their impact on neighboring genes may be less severe. Therefore, reduced selection pressure as well as other properties inherent to MITEs and SINEs could contribute to their apparent success. It is a fascinating question as to how SINEs and MITEs affect the evolution of the autonomous TEs that mobilize them.
Vertical Transmission and Horizontal Transfer of Insect TEs
As described above, the replicative ability that is responsible for the success of a TE may in some cases lead to its inactivation by generating defective copies or by activating host control mechanisms. Therefore, the ability to escape vertical inactivation by invading a new genome would greatly enhance the evolutionary success of DNA trans-posons. However, not all TEs have adopted this life cycle of invasion, amplification, senescence, and new invasion (Figure 6; see also Hartl et al., 1997; Eickbush and Malik, 2002; Robertson, 2002; Schaack et al., 2010).
Detection of horizontal transfer The occurrence of horizontal transfer can be supported in various degrees by three types of evidence (Silva et al., 2004; Schaack et al., 2010). First, detection of elements with a high level of sequence similarity in divergent taxa will offer strong support for horizontal transfer, although variable rates of sequence change should be considered. Second, detection of phylogenetic incongruence between TEs and their hosts will also provide relatively strong support for horizontal transfer. However, this alone will not be convincing, especially in light of the high levels of intra-genomic diversity of TE families observed in insects. In other words, the existence of multiple TE lineages could confound the phylogenetic analysis, as paralogous lineages may be treated as orthologous ones. Finally, horizontal transfer may be inferred when "patchy" distribution of a TE among closely related taxa is observed. This type of support is weak, as loss of TEs from sister taxa may result from a phenomenon similar to assortment of an ancestral polymorphism (Silva et al., 2004).
Horizontal transfer and vertical transmission in insects: differences between different groups of TEs The first case of eukaryotic horizontal transfer was reported in Drosophila, where the P element was shown to have invaded the D. melanogaster genome during the last century from a species in the D. willistoni group (reviewed in Kidwell, 1992). Evidence for this lateral event includes all three types of support described above, and is therefore widely accepted (Silva et al., 2004). Further analyses of a large number of P element sequences from many Drosophila species showed that many horizontal transfer events must have occurred to account for the current distribution pattern of P in Drosophila (Silva and Kidwell, 2000; Silva et al., 2004). Another spectacular case of horizontal transfer involves the mariner transposons, which was also first discovered in insects. The mariner transposon family has been implicated in hundreds or more horizontal transfer events among a wide range of animal species, including a large number of insects across different orders (Robertson, 1993, 2002). Other examples of horizontal transfer of DNA transposons, which involve insects, include ITmD37E, Harrow, Helitrons, hobo, Hosimary, and piggyBac.Horizontal transfer has also been shown for LTR retrotransposons in insects. An earlier example involves the Drosophila copia element (Jordan et al., 1999). Copia elements in D. melanogaster and D. willistoni, two divergent species that separated more than 40 million years ago, showed less than 1% nucleotide difference. It appears that copia jumped from D. melanogaster to D. willistoni. Horizontal transfer of the Drosophila gypsy element has also been reported (Terzian et al., 2000; Vazquez-Manrique et al., 2000). Evidence for horizontal transfer of LTR retrotransposons is accumulating (e.g., Kotnova et al., 2007; Cordeiro et al., 2008). As described in section 3.4.4, comprehensive analysis identified over 100 potential horizontal transfer events by more than 20 TEs among the 12 Drosophila species (Loreto et al., 2008; Bartolome et al., 2009). Schaack and collegues (2010) counted 67 horizontal transfer events involving LTR retrotransposons, 58 horizontal transfer events involving DNA transposons, and only 3 events involving non-LTR elements in Drosophila.
Possible reasons for differing propensities for horizontal transfer Two reasons have been proposed to explain the apparent differences in the prevalence of horizontal transfer events between DNA transposons and non-LTR retrotransposons (Eickbush and Malik, 2002, and herein). The first is that DNA transposons need horizontal transfer for their long-term survival, but non-LTR retrotransposons appear not to be dependent on such rare events in evolution. Defective copies of DNA transposons retain the ability to be transposed as long as they have the cis-acting signals such as the TIRs. This indiscrimination leads to the inevitable fate of inactivation of the entire transposon family. Therefore, horizontal transfer offers a much-needed escape from the above vertical inactivation, which greatly enhances the evolutionary success of DNA transposons. On the other hand, it has been shown that the reverse transcriptase of non-LTR retrotransposons tends to associate with the mRNA molecules from which they were translated (Wei et al., 2001). This cis-preference would bias transposition events in favor of the active elements, thus providing a mechanism to longer sustain the non-LTR retrotransposons. However, the cis-preference is not enough to prevent the highly successful retrotransposition of SINEs in insects (Adams et al., 1986; Tu, 1999) and other organisms (Lander et al., 2001), which presumably borrow the retrotransposition machinery from non-LTR retrotransposons. It will be interesting to see how SINEs affect the evolution of their non-LTR retrotransposon "partners." The second explanation is that the transposition process of non-LTR retrotransposons does not involve an extrachromosomal DNA intermediate, which may be important in horizontal transfer (Eickbush and Malik, 2002; see also below). In addition, DNA transposons use their transposase for integration, while non-LTR retrotransposons require more extensive involvement of host repair machinery. If the repair machinery involved in retrotransposition is species-specific, the activity of a non-LTR retrotransposon may be more restricted to its original host. Therefore, DNA transposons may be more predisposed to horizontal transfer than non-LTR retrotransposons. LTR retrotransposons form an extrachromosomal DNA intermediate, and use transposase-like integrase for integration. Therefore, LTR retrotransposons have access to the same horizontal transfer mechanisms as the DNA transposons, although their life cycle may not require horizontal transfer because defective copies are thought not to be a major factor (Eickbush and Malik, 2002). In addition, some LTR elements may acquire env protein and convert to infectious viruses that could be transmitted horizontally. It should be noted that the above are general statements, and the propensity to horizontal transfer may vary among individual families within the three groups discussed here.
Mechanisms of horizontal transfer Mechanisms of horizontal transfer are poorly understood, although direct transfer of the extrachromosomal DNA intermediate and indirect transfer through a viral vector have been proposed as possible mechanisms (Eickbush and Malik, 2002; Silva et al, 2004). Geographical and temporal overlap between the donor and recipient host species may be essential. An intriguing case of horizontal transfer of a mariner element between a parasitoid wasp and its lepidopteran host offers a good example of such overlap (Yoshiyama et al., 2001). Similarly, four TEs discovered in the kissing bug Rodnius prolixus were shown to be nearly identical to those found in the bug’s vertebrate hosts.The intracellular symbiont Wolbachia has also been suggested as one possible vector for horizontal transfer (reviewed in Schaack et al., 2010).
As the genome sequences rapidly accumulate, the numbers of examples of horizontal transfer are likely to increase. With the rapid progress in high-throughput sequencing technology and the continuing reduction of sequencing cost, it is perhaps time to move beyond "accidental" discovery of horizontal transfer and perform broad low-coverage genome sequencing surveys of many species with ecological overlap. Such surveys will identify repetitive sequences that show unexpectedly high identity between species, which will lead to candidates of very recent horizontal transfer events. This in turn will likely offer opportunities to investigate the mechanisms and circumstances of horizontal transfer, because the factors required for such lateral transfer may still be accessible for examination.
Other Possible Evolutionary Strategies
In addition to horizontal transfer and vertical extinction, recent studies suggest that there might be a third way, or an alternative strategy, which may be adopted by some TEs (Lampe et al., 2001). On the basis of the loss of interaction between mariner transposons of slightly changed TIRs, it was proposed that intra-specific or intra-genomic diversification of mariner transposons may allow the newly diverged mariner to start a new lineage. Although this requires the coevolutionary events to occur in both the transposase and the TIRs, this scenario would provide the transposon the opportunity to escape vertical inactivation, because it is now virtually a brand new element in a virgin genome owing to the loss of interaction between itself and its relatives in the genome. Genome sequencing has provided increasing opportunities to survey the diversity of different families of TEs. Our recent analysis showed a large number of lineages of non-LTR retrotransposons of the CR1 and Jockey clades in An. gambiae (Biedler and Tu, 2003). Given the presence of multiple recently active lineages within the CR1 and Jockey clades, it is tempting to speculate that the observed diversity may be driven by competition among different non-LTR families, or by attempts to escape suppressive mechanisms imposed by the host.
On the other hand, some TEs are recruited for host functions, and thus become "domesticated" (Lander et al., 2001). This type of molecular domestication is the ultimate case of trading "freedom" for "security." It allows TEs to sustain and positively impact the host, examples of which will be discussed in section 3.8. Strictly speaking, these domesticated TEs are no longer TEs. However, it is theoretically possible that these "domesticated" TEs could revert back to their "old ways" on rare occasions.
Understanding the Intra-Genomic Diversity of Insect TEs
High levels of TE diversity have been reported in the many insect genomes (see, for example, Holt et al., 2002; Kaminker et al., 2002; Kapitonov and Jurka, 2003a; Nene et al., 2007; Arensburger et al., 2010). The evolutionary process that generated this diversity may also be quite variable. It is possible that the evolution of some TEs may be a complex mix of both vertical transmission and horizontal transfer events. Parsing out the results of intra-genomic diversification from those of horizontal transfer events may require additional data from related species. Understanding the process responsible for the intra-genomic diversity of insect TEs and the potential interactions between different TE families in insect genomes will be both challenging and rewarding. A summary of the current hypothesis on TE evolution is illustrated in Figure 6. Some aspects of this model are better suited for DNA transposons that have a high propensity for horizontal transfer.
TEs in Insect Populations
Fundamental Questions and Practical Relevance
In general, the increase of TE copy number through transposition is balanced by selective forces against the potential genetic load of TEs on host fitness (Nuzhdin, 1999). The previous section discussed the control of TE transposition rate and other mechanisms to minimize their deleterious effects. This section attempts to describe the population dynamics affecting the spread of TEs in insect genomes. Earlier work on TEs in Drosophila populations suggest that the copy numbers in euchromatic regions are low, and most euchromatic copies exist at very low frequency (< 5%) in the population (summarized in Petrov et al., 2003). This was interpreted as evidence of selection against individual TE copies in nature (Charlesworth and Langley, 1989). It has been hypothesized that this selection is against three types of potentially overlapping deleterious effects by TEs, including insertional muta-genesis, transcriptional or translational cost, and ectopic recombinations between similar copies of TEs in different chromosomal regions (Charlesworth and Langley, 1989; Nuzhdin, 1999; Kidwell and Lisch, 2001; Petrov et al., 2003). For quite some time, one of the major questions in this field has been parsing out the main factors containing the spread of TEs in natural populations. However, recent analysis showed that many Drosophila TEs have high population frequencies and provide adaptive roles during evolution (reviewed in Gonzalez and Petrov, 2009). From an applied perspective, TEs have been proposed as tools to genetically drive the spread of beneficial genes through insect populations to control infectious diseases (e.g., Ashburner et al., 1998; Alphey et al., 2002). For such a sophisticated approach to work, it is important to understand the population dynamics of TEs in their insect hosts.
Experimental Approaches
In situ hybridization of the Drosophila polytene chromosomes was the main workhorse in early studies of the population dynamic of different TE families (Charles-worth and Langley, 1989). Although extremely useful, this method only works in species with accessible polytene chromosomes, and it is not efficient in detecting short regions of sequence similarities to the probe (Petrov et al., 2003). An alternative method, genomic Southern blotting, was also used to study TE insertions and excisions in Drosophila (Maside et al., 2001). Although Southern blotting is a good method to estimate TE copy numbers, it is not reliable when the numbers are high. In addition, it may not be able to detect low-frequency sites in cases where multiple small insects have to be pooled to obtain enough high-quality genomic DNA. However, a single Drosophila can provide enough DNA for a Southern experiment.
TE display, a genome-scale detection method for TE insertions, has been used recently in Drosophila and mosquitoes (Biedler et al., 2003; Yang and Nuzhdin, 2003; Bonin et al., 2008; Subramanian et al., 2008). Because it is a PCR-based method, TE display allows investigation of multiple TE families using genomic DNA isolated from individual insects. Although extremely powerful, this method cannot reliably distinguish homozygous from heterozygous insertions. However, TE display will allow recovery of a specific insertion site by sequencing the corresponding band. The sequence flanking the TE copy can be used to locate the specific site by searching the genome database, if available, or by inverse PCR. Therefore, sequences flanking a TE at a specific locus can be used as primers to amplify genomic DNA isolated from an individual sample (Figure 5). When the PCR products are run on an agarose gel, individuals with insertions at both alleles will show a single high molecular mass band, while individuals with no insertions at either allele will give a single low molecular mass band (Figure 5). Individuals that have heterozygous alleles will give both bands. Thus, this locus-specific approach is co-dominant. When the genome sequence or a bulk of genomic sequences are available for an insect species, it is not absolutely necessary to couple locus-specific PCR with TE display, because a number of TE insertion sites will already have been available for analysis (Petrov et al., 2003). However, TE display can facilitate the investigation by providing a rapid scan of a large number of loci, and by providing initial assessment of the level of polymorphism.
Recent Advances
Population studies mainly of Drosophila TEs suggest that selection against ectopic recombination may be one of the major factors in containing TE copy number in nature (Bartolome et al., 2002; Kidwell and Lisch, 2001; Rizzon et al., 2002; Petrov et al., 2003). The previously mentioned small RNA pathways may also be a critical factor controlling TE copy numbers (Lee and Langley, 2010). Recent surveys in D. melanogaster revealed an unexpected number of sites that are either fixed or of high frequency (Petrov et al., 2003; Yang and Nuzhdin, 2003), and a number of examples of TE insertions providing adaptive roles have been discovered (e.g., Maside et al., 2001; Petrov et al, 2003; Lerman and Feder, 2005; Gonzalez et al., 2008, 2010; Gonzalez and Petrov, 2009). Much more will be learned as genome sequences from individual insects within natural populations become available.
It may reasonable to assume that TE copy number and insertion frequency in a population are highly dynamic parameters that can be influenced by TE-spe-cific factors such as its intrinsic ability for transposition and self-regulation, by species- or genome-specific factors such as deletion and recombination rate and genomic control of transposition, and of course by effective population size and other ecological factors. As most population surveys only reflect a cross-section during evolution, the relative stage of a TE in its life cycle (Figure 6) is also an important consideration (Vieira et al., 1999).
