This is a proposal to determine the generality and significance of our finding that, in striking contrast to all other eukaryotes that we and others have examined, rotifers of the Class Bdelloidea appear not to contain retrotransposons. Confirmation of the absence of retrotransposons and, possibly, of DNA transposons as well, would represent an extraordinary departure from the universal occurrence of such elements in all other eukaryotes and would therefore have the potential of providing new insights into the evolution, maintenance, and role of mobile genetic elements. Moreover, such a finding would be of fundamental interest because of another highly unusual and possibly related attribute of bdelloid rotifers, their apparent asexuality.

Contrary to the generalization that the loss of sexual reproduction is a dead end in evolution, the entire Class Bdelloidea appears to have evolved for tens of millions of years without males, meiosis, or any form of sexual reproduction. A finding that bdelloids lack RNA and/or DNA transposons would therefore suggest that such elements are not retained in asexual lines over evolutionary time periods and/or that their loss or inactivation was a pre-condition for the success of bdelloids in avoiding extinction and undergoing extensive evolution without sexual reproduction. Thus, investigation of the status of mobile elements in this novel system may also reveal fundamental relationships between the presence of such elements and the evolution and maintenance of sexual reproduction.

We have initiated two complementary experimental approaches to be followed in the proposed research. The first utilizes polymerase chain reaction (PCR) with nested sets of degenerate primers designed to amplify specific conserved domains that are diagnostic for each of the major classes of RNA and DNA transposons. Resulting amplicons are cloned and sequenced to verify their identity. Such screening will be extended to include diverse species of bdelloid rotifers and, as controls, diverse monogonont rotifers and certain other animal phyla not previously examined for such elements. The second approach involves the sequencing of approximately a megabase of bdelloid DNA from our cosmid libraries in order to identify mobile elements for which we may have no effective PCR primers, including ancient inactive elements that may have diverged beyond the detection capability of PCR and other solution hybridization methods.

C.1 OBJECTIVE, SIGNIFICANCE AND RELATION TO LONGER TERM GOALS

The objective of this proposal is to test present indications that, unlike all other eukaryotes that have been examined, rotifers of the Class Bdelloidea do not have active mobile genetic elements and to detect any remnants of such elements that may once have been present. In PCR screens of genomic DNA from three different species of bdelloid rotifers we have found no indication of nucleotide sequences that code for reverse transcriptases characteristic of the major classes of retrotransposons. In contrast, using the same experimental methods, we readily find such sequences in all other eukaryotes that we have examined, including rotifers of the Class Monogononta and representatives of 19 different phyla.

The absence of retrotransposons in the Bdelloidea, and, possibly, of DNA transposons as well would constitute an intriguing exception to the prevalence of such elements throughout the animal and plant kingdoms. The significance of such a finding would lie in its potential for providing new experimentally testable insights into the evolution, maintenance, and role of mobile genetic elements.

Demonstration that bdelloid rotifers lack retrotransposons and/or DNA transposons would also be of great interest in connection with another highly unusual attribute of the this taxon, its apparent lack of sexual reproduction and genetic recombination. Since the Bdelloidea almost certainly descend from ancestors that possessed mobile elements, the absence of transposons in the Bdelloidea would suggest that they cannot be retained in asexual lines over evolutionary time periods and/or that the absence of mobile elements was a pre-condition for the success of bdelloids in avoiding extinction and evolving without sexual reproduction. As discussed in the following section, the proposed research bears directly on hypotheses for the maintenance and proliferation of mobile elements in populations and on theories for the evolution and maintenance of sexual reproduction and the extinction of asexual lines.

C2. RELATION TO PRESENT STATE OF KNOWLEDGE

Few species of animals or plants are entirely parthenogenetic, and those that are parthenogenetic rarely comprise an entire genus, let alone a taxon of higher rank (White, 1970; Bell, 1982; Richards, 1986). It is therefore thought that the complete loss of sexual reproduction is an evolutionary dead end. Against this generalization, however, the entire Class Bdelloidea of the Phylum Rotifera stands out as an apparently radical exception (Mayr, 1963; Maynard Smith, 1976).

Bdelloid rotifers are small fresh-water invertebrates of cosmopolitan distribution classified in four families, 19 genera, and more than 350 species. The individual animal has about 1,000 nuclei, with muscles, tactile and optical sensory organs, feeding structures, digestive and secretory organs, and gonads. Eggs are produced by mitosis rather than meiosis and neither males nor hermaphrodites are known within the class (Hyman, 1951; Hsu, 1956a, 1956b; Donner, 1965; Ricci, 1987; Wallace and Snell, 1991). Bdelloid rotifers comprise by far the largest animal taxon that appears to be entirely asexual.

Although these remarkable observations do not preclude rare or unrecognized forms of sexual reproduction, our laboratory has obtained molecular genetic evidence that bdelloid rotifers have indeed evolved without sexual reproduction. Our approach makes use of the fact that the genome of a sexually-reproducing diploid contains two sets of chromosomes that are kept closely homologous by segregation and genetic drift. In contrast, if reproduction is only mitotic and there is no germ-line genetic recombination of any kind, segregation cannot occur and the accumulation of mutations will cause nucleotide sequences that initially are closely homologous to diverge, so that after many millions of years individual genomes will no longer contain closely homologous chromosome pairs. In accord with this expectation, we have found highly divergent copies of the hsp82 heat shock gene in individual rotifers of four different bdelloid species representing three different families. Also, in less extensive studies of bdelloid genomes, we have found only highly diverged copies of three other genes (RNApolIII, tbp, and tpi). In contrast, as expected for sexually-reproducing diploids, we find only a pair of closely homologous copies of hsp82 in the genomes of individual rotifers of the Class Monogononta, which are facultatively sexual diploids (D. Mark Welch and M. Meselson, manuscript in preparation).

In a more direct test for the presence of homologous chromosome pairs in bdelloid rotifers, we are conducting fluorescent in situ hybridization of YAC and cosmid genomic probes to bdelloid metaphase chromosomes. The one YAC clone so far tested hybridized intensely to a single chromosome, with no evidence of hybridization elsewhere (J. Mark Welch and M. Meselson, unpublished data). While these initial hybridization results need to be extended to additional clones, the lack of males and meiosis, the unusual presence of highly diverged copies of each gene examined, and, most recently, our preliminary cytogenetic results all support the view that the Class Bdelloidea has evolved without sexual reproduction.

If bdelloid rotifers have indeed evolved without sexual reproduction, the question arises as to what characteristics of the genome, in addition to the absence of closely homologous chromosomes, may be associated with ancient asexuality. A characteristic we have chosen to examine is the content of mobile genetic elements and, in particular, the possibility that, in contrast to sexually-reproducing species, rotifers of the Class Bdelloidea lack such elements.

Although the initial impetus for testing bdelloid rotifers for the presence of transposons derived from speculation regarding the possible role of such elements in the evolutionary maintenance of sex, our present evidence that bdelloids lack retrotransposons is reason enough to pursue this highly atypical finding in a number of different directions: (i) extension of PCR screens to additional classes of retrotranposons and to DNA-based transposons; (ii) Southern blot comparisons of bdelloid and monogonont genomic DNA to detect possible global differences in the content of repeated elements and to isolate any such elements that may be present in bdelloids; (iii) examination of additional bdelloid and monogonont species of rotifers; (iv) sequencing of several thousand kb of bdelloid genomic DNA to detect and characterize possible active or relict mobile elements undetectable by PCR.

As stated above, the proposed research bears on specific questions regarding the maintenance and proliferation of mobile elements and on hypotheses for the possible role of such elements in the maintenance of sexual reproduction. In particular, if further work confirms present indications that bdelloid rotifers lack active mobile genetic elements, attention would be focused on two different theoretical explanations for why such elements may not be present in ancient asexual species.

First, the elimination of syngamy as the principal vehicle for horizontal proliferation of mobile elements may lead to their eventual loss in asexual lines. In addition to changes brought about by genetic recombination and segregation, the population of active transposons in the genome of a sexually-reproducing species is subject to depletion by mutation or excision of individual elements and to replenishment by autonomous replication and by horizontal transfer during sexual reproduction or, more rarely, by asexual transfer from the same or a different species (Charlesworth et al., 1994; Kidwell, 1993). If sexual reproduction and recombination are lost, most horizontal input ceases and the population of active transposons may decrease and perhaps eventually reach zero. Once lost in an asexual line, a given transposon type can be regained only by asexual transmission, which may be exceedingly infrequent.

Second, while the above argument envisages the loss of mobile elements as a consequence of the loss of sexual reproduction, a quite different line of thought would suggest that the loss of such elements may be a precondition for long-term survival and evolutionary success of asexual taxa. This possibility is suggested by hypotheses that ascribe to genetic recombination a principal role in limiting the accumulation of deleterious mutations in sexually-reproducing populations by giving rise to segregants that harbor fewer deleterious mutations than either of the two recombining parental genomes. According to these hypotheses, whether by stochastic (Muller, 1963) or deterministic (Crow, 1994; Kondrashov, 1994) processes, asexual lines should lose out to sexual ones and ancient species-rich asexual taxa, such as the Bdelloidea appear to be, should not exist.

On this view, however, the extinction of asexual lines could be avoided, or at least greatly delayed, by sufficiently reducing the accumulation of deleterious mutations, thereby allowing time for species radiation and the formation of higher-rank taxa. We have proposed that if the accumulation of deleterious mutations is indeed a major cause of the early extinction of asexual lines, it is the proliferation of mobile elements, rather than the accumulation of simple sequence changes, that contributes most of the relevant deleterious load⁽¹⁾ (Arkhipova et al., 1995). In Drosophila, the great majority of spontaneous visible mutations is caused by transposon insertion (Green, 1988). Thus, the inactivation or loss of mobile elements in an early ancestor of modern bdelloids, either before or after the loss of sexual reproduction, could have averted the early extinction usually suffered by asexuals, allowing bdelloids to evolve extensively without sexual reproduction. On this hypothesis, such elements should be lacking in bdelloid rotifers.

If, contrary to our preliminary findings, active mobile elements are found in bdelloids and since our studies of various bdelloid species indicate that they experience nucleotide substitutions with normal frequency, doubt would be cast on all hypotheses for the maintenance of sexual reproduction and the long-term inviability of asexuals that depend on mutation reduction (reviewed in Kondrashov, 1994).

In any case, we believe that the possibility, suggested by the preliminary work described below that bdelloid rotifers present a radical exception to the general occurrence of mobile genetic elements in eukaryotic organisms, should not be left unexplored.

C.3 PRELIMINARY RESULTS

Presently, the sole source of outside support for our laboratory is NSF grant MCB-9514279 "Molecular Genetic Studies of Bdelloid Rotifers". It has provided support for our studies of sequence and chromosome divergence in bdelloid rotifers, as described in section C.2 above and a renewal is pending. However, neither of these grants include transposon studies, for which we have no outside support. The preliminary studies of transposons described below were funded from a limited Harvard University research account.

All known eukaryotic transposons can be divided into two classes: (I) retrotransposons, which transpose via an RNA intermediate synthesized by the element-encoded reverse transcriptase, and (II) DNA-transposons, which transpose as DNA by a cut-and-paste mechanism using the element-encoded transposase (for review see Arkhipova et al., 1995). Both classes are subdivided into superfamilies, in which the element-encoded enzymes exhibit more sequence similarity to other members within the same superfamily, even in very distantly related host species, than to members of a different superfamily, even within the same host.

The three superfamilies of retrotransposons are called copia/Ty1, gypsy/Ty3, and LINE/non-LTR retrotransposons. For DNA transposons, the two superfamilies are called Tc1/mariner and hobo/Ac/Tam (hAT) (Hartl, 1989; Robertson, 1993; Calvi et al., 1991). Non-autonomous mobile elements, which do not themselves encode functional reverse transcriptases or transposases, usually co-exist in a given genome with full-length autonomous transposons (or, in some cases, incomplete elements) that provide the transposition machinery.

Transposon sequences can be detected in eukaryotic genomes as components of middle repetitive DNA, as insertions into various genes, or as sequences having homology to previously known transposons. A way that has proven effective to screen for certain transposons is to employ a single step of PCR amplification with degenerate primers spanning the most conserved domains of reverse transcriptases or transposases. This approach has successfully been used to amplify those transposon sequences which contain at least two six-amino-acid blocks of similarity to a given query transposon-encoded protein (Flavell et al., 1992; Voytas et al., 1992; Wichman and Van Den Bussche, 1992; Robertson, 1993). A problem with this approach lies in the relative shortness of the blocks of conserved amino acids in many reverse transcriptases and transposases, making the detection of distantly related transposons uncertain. Reverse transcriptases contain seven characteristic "signature" domains distributed over more than a kilobase with variable spacings (Figure 1). However, each domain displays only two or three amino acids that are well-conserved across the entire superfamily (Poch et al., 1989; Xiong and Eickbush, 1990). In DNA mediated transposons, several analogous short conserved blocks of homology exist in most transposases (Doak et al., 1994).

To compensate for the lack of extensive conservation in reverse transcriptases, we have successfully used two-step PCR amplification with nested primers (for general reference, see White, 1997). The procedure we have designed takes advantage of the most conserved reverse transcriptase domains, designated A, B, C, and E (Figure 1). The first set of highly degenerate primers we employed, targeted to domains A and E (DhXX[A/G][F/Y] and [Y/F]LGXXh), can hybridize to 1-2x10⁶ potential target sequences. The degeneracy of the primers was reduced to 36,000-fold by inclusion of several deoxyinosines. The amplification products obtained with these primers, substantially enriched in retrotransposon-related sequences, were subjected to a second round of amplification using 1-2x10⁴-fold degenerate primers against domains B and C, located between domains A and E. Optimization of annealing temperatures, numbers of cycles, and buffer composition was done with dilute samples of known cloned reverse transcriptase sequences for each primer combination. This two-step procedure reproducibly yields specific PCR products of characteristic length (ca. 120 bp) which are easily visualized on polyacrylamide gels (Figure 2). Using the same A-E primer pools for the first amplification and the appropriately designed B-C primer pools for the gypsy-like and LINE-like superfamilies for the second amplification, we have applied this procedure to a wide variety of species (Figure 2).

The diagnostic bands corresponding to gypsy-like retrotransposons (Figure 2, panel A) and those corresponding to LINE-like retrotransposons (Figure 2, panel B) are both undetectable in genomic DNA isolated from three species of bdelloid rotifers (Philodina roseola, Habrotrocha constricta, and Adineta vaga). In contrast, as seen in Figure 2, such bands are clearly evident under the same experimental conditions in amplifications of DNA from a monogonont rotifer (Brachionus plicatilis) and from all other animal species we have analyzed, including echinoderms, insects, amphibians, fishes, and mammals. A band diagnostic for gypsy-like retrotransposons, presumably originating from the yeast element Ty3, is also detectable in DNA from Saccharomyces cerevisiae. As expected, there is no indication of the presence of LINE-like elements in this species, the entire genome of which is known to be devoid of such elements (Cherry et al., 1997). Note that the bands corresponding to LINE-like elements from different species appear more heterogeneous in length than those corresponding to gypsy-like elements, in agreement with the more variable distance between domains B and C in LINE-like reverse transcriptases (Xiong and Eickbush, 1990).

The validity of the double-PCR method was demonstrated by cloning and sequencing six different clones of presumptive gypsy-like amplicons and five different clones of candidate LINE-like amplicons from a total of nine different species, including two species of monogonont rotifers. As may be seen in Figure 3, all eleven clones had significant homology to known classes of retrotransposons at the amino acid level, as determined by the BLAST search algorithm (Altschul et al., 1997). Moreover, five of these clones also had significant homology at the nucleotide sequence level to known retrotransposons. One of the D. melanogaster clones was identical to the known D. melanogaster LINE-like retrotransposon BS, except for a 9-bp deletion. Another D. melanogaster clone had 75% homology to the gypsy retrotransposon. A clone from D. virilis was 66% identical to the known gypsy-like micropia retrotransposon of D. hydei. Two cloned amplicons from C. elegans were 100% and 94% identical to LINE-like reverse transcriptase sequences contained in the C. elegans genome sequencing project data base. Figure 3 shows the corresponding amino acid sequence alignments for eleven of the cloned amplicons, compared with reverse transcriptase sequences of known retroelements.

It should be noted that while three of the gypsy-like and four of the LINE-like clones contained intact open reading frames, four of the gypsy-like and one LINE-like sequence were interrupted by a frameshift and/or a stop codon. Thus, the technique allows one to visualize a mixture of active and inactive transposons present in genomic DNA.

Several segments of DNA from three different species of bdelloids were cloned from the region of the gel that would correspond to the size expected for retrotransposon amplicons. They were sequenced and found to have no homology to transposons or to any other sequences in GenBank. This negative result, together with our consistently positive findings, including inactive LINE-like and gypsy-like reverse transcriptases in monogononts and all other eukaryotes that we have examined, indicates that bdelloid rotifers have lacked the corresponding retrotransposons for a very long time.

Initial attempts have been made to develop PCR primers for detecting DNA transposons of the Tc1/mariner superfamily. Previous studies employing only a single PCR amplification, while producing positive results in many species, failed to detect mariner-like sequences in a substantial proportion of phyla tested (Robertson, 1997). In our experiments, we find that improved amplification results are obtained by conducting two steps of amplification, using two different nested sets of primers. Although there are only three conserved domains in the tripartite DD(35)E motif of the transposase gene (Doak et al., 1994), the most extensive region of homology, a nine-amino-acid motif near the C-terminus, is sufficiently long to accommodate two different primer sequences. This allowed us to design two nested primer sets for mariner-like transposases. Two rounds of amplification have yielded intense positive signals in representatives of four phyla tested, and their identity as DNA transposons was confirmed by cloning and sequencing a PCR amplicon in one case, the planarian Dugesia tigrina. Additional primers will be designed in order to amplify Tc1-like transposases. Once proven effective in amplifying members of the Tc1/mariner superfamily, these primer sets will be employed for the screening of genomic DNA of various species of bdelloid and monogonont rotifers.

The highly unusual result of failing to detect reverse transcriptase-related sequences by amplification calls for a more extensive examination of bdelloid rotifer genomes for the presence of active retrotransposons and the extension of the analysis to include DNA transposons. Moreover, the success of our procedure in detecting retrotransposons in diverse taxa representing more than half of the major phyla makes it desirable to extend our survey to include the remaining major phyla, thereby further testing the generally held belief in the ubiquity of transposons and significantly extending our knowledge of their distribution.

C.4 GENERAL PLAN OF WORK TO BE UNDERTAKEN

Nested PCR analysis

DNA from four different species of bdelloid rotifers, representing three of the four families of the Class Bdelloidea (Philodina roseola and Macrotrachela quadricornifera [Philodinidae]; Habrotrocha constricta [Habrotrochidae]; and Adineta vaga [Adinetidae]) will be examined by nested primer PCR for the presence of DNA transposons and for all known retrotransposon superfamilies, including LINE, gypsy/Ty3, copia/Ty1 and Pao/Tas/BEL (Xiong et al., 1993; Felder et al., 1994). Cultures of these bdelloid rotifers are maintained in our laboratory. They have been started from single eggs and are fed E. coli to avoid contamination with other eukaryotic organisms. The fourth bdelloid family, the Philodinavidae, is extremely rare but we will attempt to obtain a representative of it from a colleague at the University of Milan, Dr. Claudia Ricci. DNA from four species of monogonont rotifers (Brachionus plicatilis, Brachionus calyciflorus, Sinantherina socialis, and Eosphora ehrenbergii) representing three different families (Brachionidae, Flosculariidae and Notommatidae) will also be examined, using the same experimental procedures.

In addition to the primer pools used in the preliminary studies described above, we will design new nested primer sets for amplification of segments bounded by conserved domains of reverse transcriptases and transposases of the remaining RNA and DNA transposon superfamilies. A full compilation of all reverse transcriptase and transposase sequences in current genetic data bases will provide the basis for generating refined superfamily-specific consensus sequences to be used for primer design and for determining inter-domain distances that specify amplicon lengths. As an example of this procedure, Figure 4 shows a partial lineup for gypsy-like transposons, based on a computer search of the available data. Similar alignments can be generated for every transposon superfamily.

Transposons with reverse transcriptase and integrase domains deviating substantially from consensus sequences have occasionally been found in certain taxonomic groups, including fungi and protozoans. Although these might escape detection, this possibility can be reduced by using primers for several primary amplification domains simultaneously. For example, if any two oppositely directed primers bind sufficiently to the target sequences, a combined use of all four reverse transcriptase outer domains 1, 2, D, and E for primary amplification of LINE-like retrotransposons (Figure 1) should produce enough enrichment for the second round of amplification with primer sets corresponding to combinations of the most conserved inner domains A, B, and C. The existence of the AB, AC, or BC band would provide evidence for the presence of LINE-like transposons, subject to confirmation by sequencing of resulting amplicons. In order to complete an exhaustive search, these techniques will be applied to other known transposon superfamilies.

In order to monitor the quality of DNA preparations used for PCR so as to preclude false negatives, parallel amplifications will be carried out with sets of degenerate primers for highly conserved genes that we have successfully used in other work (hsp82, tbp, tpi, and RNA polIII). In fact, problems with DNA quality have been encountered only rarely.

Cloning and sequencing of candidate transposons

Amplicons from major bands presumably corresponding to reverse transcriptases or transposases will be cloned in bacterial plasmids (pBluescript II SK(+), Stratagene Corp.) and sequenced by the dideoxy method (fmol DNA Cycle Sequencing System, Promega Corp.) in order to verify their transposon nature. Special emphasis will be placed on obtaining transposons from invertebrate species that are relatively closely related to bdelloid rotifers (diverse monogonont rotifers, Acanthocephala, and Gastrotricha) to verify that these taxa do contain active transposons. The cloned PCR fragments of transposon origin will be used as probes for screening the corresponding genomic libraries in order to obtain full-length transposon copies. For these particular taxa, sequence analysis of such copies will demonstrate whether they contain intact open reading frames, target site duplications, long terminal repeats and other hallmarks characteristic of active or recently immobilized transposons. Cloned transposon probes will also be used in high-stringency Southern blotting experiments to estimate their copy number in the corresponding genomic DNAs and to control against contamination by foreign DNA.

Should any products indicative of transposons appear in PCR analyses of bdelloid rotifer DNA, they will be cloned and sequenced in order to reveal possible relationships to transposons. If any such relationship is seen, the corresponding clone will be used as a probe in high-stringency Southern analyses of genomic DNA to verify that the amplified segment is not a contaminant. Any such clones will also be used to probe our bdelloid genomic libraries and the resulting clones will be sequenced to detect any further similarity to transposons.

Comparison of middle repetitive DNA content

in bdelloid and monogonont rotifers

Having already found reverse transcriptases characteristic of specific families of retrotransposons in monogononts and in diverse species of other phyla but not in bdelloids, it would be of interest to compare the overall pattern of middle repetitive DNA distribution in bdelloids with that in monogononts. To this end, we will first carry out Southern blot hybridization analysis, probing restricted bdelloid and monogonont genomic DNA with the corresponding labeled genomic DNA.

While this approach will allow us to visualize any major differences in the amount and diversity of repeated sequences, we will also screen our P. roseola cosmid library with radiolabeled genomic DNA in order to select specific clones containing sequences that may be repetitive in the genome. Clones containing ribosomal RNA sequences but no other repetitive elements will be excluded from the screen by the addition of an excess of non-radioactive DNA from P. roseola cDNA clones of rDNA genes (available from Dr. Elizabeth Walsh). Cosmids which yield relatively intense hybridization signals upon such screening will be further analyzed by restriction enzyme digestion and Southern blotting against radiolabeled genomic DNA to identify distinct classes of repetitive DNA that may be present. Any restriction fragments of interest will then be subcloned and sequenced to identify transposon-like sequences that may be present but which were not detected by our PCR analyses. Furthermore, five to seven cosmids suspected to carry transposons will be selected for large-scale sequencing (see below), in order to include a representative fraction of cosmids carrying moderately repetitive DNA into the sequencing project.

Large scale DNA sequence analysis

In order to detect possible relict transposons in bdelloid rotifers that have diverged to the point that they cannot be found by PCR amplification, we will sequence a number of cosmids from our bdelloid libraries in order to conduct a computer search for significant homology to known reverse transcriptases and transposases.

As may be estimated from the ability of our PCR screen of Saccharomyces cerevisiae to detect gypsy-like elements, which comprise less than 0.5% of the total genomic DNA of this yeast (Hani and Feldman 1998), our PCR methods appear capable of detecting as few as one element of nominal 5 kb length per 1,000 kb of DNA. Thus, a substantial sequencing effort would be required to reach this sensitivity. Although direct sequencing would therefore not be preferable to PCR analysis for detecting well-conserved transposons of known classes and is certainly prohibitive for the wide screen we propose to conduct in diverse phyla (see below), sequence analysis appears to be the only suitable method for screening bdelloid genomes for elements whose homology to our primers is below the level required for PCR amplification.

In numerous organisms, transposons usually comprise 5-20% of the total genome (Berg and Howe, 1989). For a 5 kb element, this corresponds to an average density of approximately one copy per 50-100 kb. In order to obtain more specific information, we have conducted a computer analysis of the recently completed 97 Mb Caenorhabditis elegans genomic DNA sequence (The C. elegans Sequencing Consortium, 1998). This reveals slightly more than 200 sequences with significant amino acid homology to LINE-like reverse transcriptases, corresponding to an average density of approximately two such elements per megabase. We find a similar genomic density for transposases of the Tc1/mariner superfamily. Since these two classes comprise a majority of the transposons in this species, this means that, on average, one out of four or five cosmids in this species contains a transposable element detectable by computer analysis. However, nearly one-half of the LINE-like transposons detected by this search are apparently inactive, as judged by the presence of stop codons and frameshifts. Moreover, they usually contain one or more substitutions in the conserved motifs of reverse transcriptase, making them undetectable by our PCR methods. If present in bdelloids at comparable frequency, such relict transposons could be identified by sequencing 10-15 cosmids. As we have already shown that bdelloid PCR is unable to detect the major classes of retrotransposons in bdelloid rotifers, the only way to detect possible highly diverged relict transposons that may remain in their genomes would be sequence analysis. The effort is worth undertaking because the detection of relict retrotransposons in bdelloids may provide information both as to how long ago they ceased to be active and as to the nature of the lesions by which they were inactivated.

Cosmid sequences, determined from subcloned segments submitted to the departmental automatic sequencing facility, will be examined by BLAST searches to detect homology to known transposons and by diagonal dot-matrix comparison with varying parameters to identify closely or distantly related repeated sequences, possibly including short non-autonomous repeats, such as solo LTRs, SINEs, or MITEs (Bureau et al., 1996; Okada et al., 1997).

Species representing only seven out of the more than 30 recognized animal phyla are reported to have been examined for transposons, one or more transposon types having been found in each case. It is therefore of interest to examine these phyla for the presence of transposons. As seen in Table 1, we have obtained and prepared DNA from representatives of 19 of the 31 major animal phyla and have analysed most of these species for LINE and gypsy/Ty3 retrotransposon superfamilies by PCR. Species from the remaining 12 phyla will be obtained from commercial suppliers (Carolina Biological Supply, Gulf Specimen Marine Laboratory, Woods Hole Oceanographic Institution) or provided by individuals listed in the Table. DNA samples will be prepared by SDS/proteinase K treatment and phenol-chloroform extraction, using thoroughly washed whole animals for microscopic organisms and dissected tissues (preferably gonads) for macroscopic ones. Although contamination with parasitic or endosymbiotic microorganisms cannot be completely excluded, the technique provides a substantial advantage to multicopy transposons, usually allowing them to be visualized well before single-copy sequences or trace contaminants are amplified to any significant extent. Moreover, as explained above, questionable cases will be further tested by Southern blot analysis as a control against contamination by foreign DNA.

Genome sequencing projects in diverse model organisms (S. cerevisiae, D. melanogaster, C. elegans, A. thaliana) have not revealed any previously unknown transposon classes, providing confidence that the primer sets we will employ will be capable of detecting whatever transposons may be present. Therefore, the wide transposon screen that we propose will provide a stringent test of the possibility that bdelloid rotifers lack transposable genetic elements. Finally, if no transposons are found in bdelloids, similar screens might be undertaken in several species suspected of being ancient asexuals but for which there is little or no supporting molecular evidence (Judson and Normark, 1996).

LDLKSGY CRLPFGL VYVDDVII YLGFIV D.melanogaster gypsy

LDLKSAF CRLPFGL VYVDDVII YLGFMV D.melanogaster burdock

IDLKSGF TRLPFGL VYIDDIII FLGFIV C.capitata yoyo

IDLAKGF LRMPFGL VYLDDIII FLGHIV D.melanogaster 297

IDLARGF VRMPFGL VYLDDMII FLGHIV D.ananassae tom

LDLKSAY LVMPFGL CYLDDILI YLGHEI F.oxysporum skippy

LDLASGF LRMPMGL VYLDDIIV YLGHII Trichoplusia ni TED

LDLHSGY TVMPFGL VYLDDILI FLGYSI S.cerevisiae Ty3

IDLAKGF LRMPFGL VYLDDIIV FLGHVL D.melanogaster 17.6

MDAFSGY KVMPFGL VYIDDMLV FLGYIV A.thaliana V MXA21

MDLRNGY QVMPFGL CYLDDILV FLGFTV B.cinerea Boty

LDLKNGF LRMPFGL VYMDDILL YLGFII D.melanogaster mdg3

LDHRSGY RVMPFGL VFLDDVLI FLGHII Z.mays Reina

IDLRSSY QVLPFGL NYIDDLLV FLGHIL T.castaneum Woot

IDLTSGY VVMPFGL IFIDDILV FLGHIV A.thaliana V BAC T32N15

LDLKSAY LVMPYGI CYMDDILI FIGYHI S.pombe Tf1

MDLRSAY KRLPFGV AYLDDITV FLGYRI Ciona intestinalis Cir1

IDLRSGY VVMSFGL VFIDDILI FLGHVV Lycopersicon esculentum RT

MDMANGY THLPFGL VYIDDVLI FLGHEI C.elegans cosmid R03D7

IDLRFRY LVMPFGL VFVDYVLI FLGHVV Lilium henryi del*

LDLMSGF TRLPFGL LYMDDLVV YLGHKC D.melanogaster mdg1

IDIVSAF LVMPFGL AYLDDILI FLGLLV M.grisea maggy

VDVRAAF LVCPFGL AYLDDILI YLGFIV M.grisea grh

LDLMSGF TRLPFGL LYMDDLIV FLGHKC D.melanogaster 412

LDMASGF LTMPFGL VYMDDIMI YLGYEV D.hydei micropia

MDLKDGY KVMPFGL AYLDDKIV YLGHKV D.buzzatii osvaldo

IDLRDAF LVMPMGL AYMDDILV FLGFII C.fulvum CfT-1

LDCSGAF VRMPYGL AYLDDIII YLGVLV Nasonia vitripennis NATE P16

LDCYSWY TSMSFGL AYVDDVVI LLGFIV Z.diploperennis Grande1-4

LDLMSGF TRLPYGL LYMDDLVV YLGHKC D.melanogaster P1 DS03023

LDCSGAF VRMPYGL AYLDDIII YLGVLM Nasonia vitripennis NATE P17

LDLKGAY LVMLFGL SYLDDILI FLGYTI F.oxysporum RT

FDMIAGF NVLPFGL VYVDDLLI YLGHKV C.elegans Cer1

MDCSGAF VRMPYGL AYLDDIII YLGVLV Nasonia longicornis NATE c21

MDGFFGY KVMPFGL VYVDDMIA LLGFVV Vicia faba (broad bean) RT

LDGYSGY RRIPFGL VFMDDFSV VLGHKI Vicia faba RT

LDLSQAY SRLVYGL VFYDDILI YLGFII B.mori mag

IDLKDAY RRMTFGL AYLDDVIV FLGFII C.elegans cosmid T23E7

IDLKDAY LRMTFGL VYLDDIII FLGFIV C.elegans cosmid T03F1

STLSSRS RHMPFGL VYLDDLLI YLGFIQ D.virilis Ulysses

MDMASGF AFVPDGL VYMDDIMV YLGYEV D.melanogaster micropia-Dm11

IDFRDAY VRMAFGL AYLDDLII FLGAVC C.elegans cosmid E03A3

IDLKDAY LRMTFGL VYWDDIII FLGAVS C.elegans cosmid K02A2

MDLQNGF NKAPFGF LYMDDIIV FLGHII D.melanogaster blastopia

FDMIMGY TRMPQGL QYFDDLLV FLGFHI D.discoideum Skipper

LDASSGF LRLPFGI TMMDDVIV FLGMIN Tripneustes gratilla Tgr1

LDLKSGY LVMPFGL VFFDDILV YLGHRH A.thaliana mito B

LDIKKAY KTMPFGL AYLDDLLI FLGLQI D.discoideum DIRS-1

LDLKDAF MVLPQGF QYVDDLLL YLGLIL H. sapiens HERV9

LDLKGAY LVMLFGL SYLDDILI FLGYTI F.oxysporum RT*

IDLKDAY TSLAFGL YYLDDICV YLGFSF P.blakesleeanus Prt1*

FDMKSGY KALPFGL LYLDDLLV WLGVVF Panagrellus redivivus PAT

MDLTKGY RRMPFGL VYLDDIVI FLGIVG Xenopus laevis Matryoshka

LDLRSGY FVMPFGL VFFDDILI YLGHIV Pine IFG7

+* ** +**** *+**+++ *** +

Figure 1. Domain structure of reverse transcriptases belonging to three different superfamilies of eukaryotic retrotransposons (Poch et al., 1989; Xiong and Eickbush, 1990).

Figure 2. Amplification of DNA segments corresponding to gypsy-like (A) and LINE-like (B) reverse transcriptases. Primary amplification was done with primers A and E described in the text. For secondary amplification, degenerate primer pools were: (A) XXhPFG for domain B and YhDDhh for domain C; (B) XGX[P/R]QG for domain B and [F/Y]ADDXX for domain C. Bdelloid rotifers (P. roseola, H. constricta and A. vaga) appear to lack reverse transcriptases of both classes.

Figure 3. Amino acid sequence alignments between conceptual translations of gypsy-like (A) and LINE-like (B) amplicon sequences obtained and cloned as described in the text and their two nearest matches from GenBank, identified in a BLAST search as gypsy-like and LINE-like reverse transcriptases, respectively. Gaps (-), stop codons (*) or frameshifts (#) were introduced to maximize alignment scores which include identical (|) and similar (:) amino acids. Cloned amplicons are in boldface.

Figure 4. Alignment of conserved amino acid residues within domains A and E (shown in red) and domains B and C (shown in green) of the top fifty matches to the gypsy reverse transcriptase in a GenBank TBLASTN search. Asterisks and pluses on the bottom designate conserved and chemically similar (hydrophobic) residues, respectively. The species and transposon names on the right designate representatives of different taxa (insects, nematodes, fungi, plants etc.), while the sequences appear in order of decreasing similarity scores from top to bottom.

Table 1. Representative species from the major animal phyla to be analyzed for the presence of transposons. The second column indicates classes of transposons identified in representatives of a given phylum by other researchers. The last two columns show the results of our preliminary PCR assays for two major retrotransposon subfamilies. Pluses denote the presence of a diagnostic PCR band; (S) indicates that the identity of the band was confirmed by cloning and sequencing. GSML - Gulf Specimen Marine Laboratory, Panacea, FL; CBS - Carolina Biological Supply, Burlington, NC; MBL - Marine Biological Laboratory, Woods Hole, MA.

Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D.J. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389-3402.

Arkhipova, I.R., Lyubomirskaya, N.V., and Ilyin, Y.V. (1995) Drosophila Retrotransposons. R.G. Landes Co. & Springer-Verlag.

Bell, G. (1982) The Masterpiece of Nature: the evolution and genetics of sexuality. Univ. of California Press.

Berg, D.E., and Howe, M.M., eds. (1989) Mobile DNA. American Society for Microbiology.

Bureau, T.E., Ronald, P.C., and Wessler, S.R. (1996) A computer-based systematic survey reveals the predominance of small inverted-repeat elements in wild-type rice genes. Proc. Natl. Acad. Sci. USA 93: 8524-8529.

Calvi, B.R., Hong, T.J., Findley, S.D., and Gelbart, W.M. (1991) Evidence for a common evolutionary origin of inverted repeat transposons in Drosophila and plants: hobo, Activator, and Tam3. Cell 66: 465-471.

The C. elegans Sequencing Consortium (1998) Genome sequence of the nematode C. elegans: A platform for investigating biology. Science 282: 2012-2018.

Charlesworth, B., Sniegowski, P., and Stephan, W. (1994) The evolutionary dynamics of repetitive DNA in eukaryotes. Nature 371: 215-220.

Cherry, J.M., Ball, C., Weng, S., Juvik, G., Schmidt, R., Adler, C., Dunn, B., Dwight, S., Riles, L., Mortimer, R.K., and Botstein, D. (1997) Genetic and physical maps of Saccharomyces cerevisiae. Nature 387 (6632 Suppl): 67-73.

Crow, J.F. (1994) Advantages of sexual reproduction. Dev. Genet. 15: 205-213.

Doak, T.G., Doerder, F.P., Jahn, C.L., and Herrick, G. (1994) A proposed superfamily of transposase-related genes: new members in transposon-like elements of ciliated protozoa and a common "D35E" motif. Proc. Natl. Acad. Sci. USA 91: 942-946.

Donner, J. (1966) Rotifers. Transl. by H.G.S. Wright. Frederick Warne & Co.

Felder, H., Herzceg, A., de Chastonay, Y., Aeby, P., Tobler, H., and Muller, F. (1994). Tas, a retrotransposon from the parasitic nematode Ascaris lumbricoides. Gene 149: 219-225.

Flavell, A.J., Smith, D.B., and Kumar, A. (1992) Extreme heterogeneity of Ty1-copia group retrotransposons in plants. Mol. Gen. Genet. 231: 233-242.

Green, M.M. (1988) Mobile DNA elements and spontaneous gene mutation, pp.41-50 in: Banbury report 30. Eukaryotic transposable elements as mutagenic agents (M.E. Lambert, J.F. McDonald, and I.B. Weinstein, eds.) Cold Spring Harbor laboratory.

Hani, J., and Feldmann, H. (1998) tRNA genes and retroelements in the yeast genome. Nucleic Acids Res. 26: 689-696.

Hartl, D.L. (1989) Transposable element mariner in Drosophila species. In: Mobile DNA (D.E. Berg and M.M. Howe, eds.), pp. 531-537. American Society for Microbiology.

Hickey, D.A. (1982) Selfish DNA: a sexually-transmitted nuclear parasite. Genetics 101: 519-531.

Hsu, W.S. (1956a) Oogenesis in the bdelloidea rotifer Philodina roseola (Ehrenberg). La Cellule 57: 283-296.

Hsu, W.S. (1956b) Oogenesis in Habrotrocha tridens (Milne). Biol. Bulletin 111: 364-374.

Hyman, L.H. (1951) Invertebrates, Vol. 3. McGraw Hill.

Judson, O.P., and Normark, B.B. (1996) Ancient asexual scandals. Trends Ecol. Evol. 11: 41-46.

Kidwell, M.G. (1993). Lateral transfer in natural populations of eukaryotes. Annu. Rev. Genet. 27: 235-256.

Kondrashov, A.S. (1994) Classification of hypotheses on the advantage of amphimixis. J. Hered. 84: 372-387.

Maynard Smith, J. (1978) The Evolution of Sex. Cambridge Univ. Press.

Mayr, E. (1963) Animal Species and Evolution. Belknap Press.

Okada, N., Hamada, M., Ogiwara, I., and Ohshima, K. (1997) SINEs and LINEs share common 3' sequences: a review. Gene 205: 229-243.

Poch, O., Sauvaget, I., Delarue, M., and Tordo, N. (1989) Identification of four conserved motifs among the RNA-dependent polymerase encoding elements. EMBO J. 8: 3867-74.

Ricci, C.N. (1987) Ecology of bdelloids: How to be successful. Hydrobiologia 147: 117-127.

Richards, A.J. (1986) Plant breeding systems. Allen and Unwin.

Robertson, H.M. (1993) The mariner transposable element is widespread in insects. Nature 362: 241-245.

Robertson, H.M. (1997) Multiple mariner transposons in flatworms and hydras are related to those of insects. J. Hered. 88: 195-201.

Voytas, D.F., Cummings, M.P., Konieczny, A., Ausubel, F.M., and Rodermel, S.R. (1992) copia-like retrotransposons are ubiquitous among plants. Proc. Natl. Acad. Sci. USA 89: 7124-7128.

Wallace, R.L., and Snell, T.W. (1991) Rotifers. In: Ecology and classification of North American freshwater invertebrates (J.H. Thorp and A.P. Covich, eds.). Academic Press.

White, M.J.D. (1978) Modes of Speciation. Freeman.

White, B.A., ed. (1993) PCR cloning protocols: From molecular cloning to genetic engineering. Methods in Molecular Biology, Vol. 67. Humana Press.

Wichman, H., and Van Den Bussche, R.A. (1992) In search of retrotransposons: Exploring the potential of the PCR. BioTechniques 13: 258-264.

Xiong, Y., and Eickbush, T.H. (1990) Origin and evolution of retroelements based upon their reverse transcriptase sequences. EMBO J. 9: 3353-3362.

Xiong, Y., Burke, W.D., and Eickbush, T.H. (1993) Pao, a highly divergent retrotransposable element from Bombyx mori containing long terminal repeats with tandem copies of the putative R region. Nucl. Acids Res. 21: 2117-2123.

1. ¹ This hypothesis for the role of transposons in the maintenance of sexual reproduction is to be distinguished from the proposal of Hickey (1982) that, as carriers of genes for conjugation in bacteria, transposons may have driven the origin of sexual reproduction.