Journal of Molecular Evolution

, Volume 66, Issue 6, pp 643–654 | Cite as

The Association Between Breeding System and Transposable Element Dynamics in Daphnia Pulex

Article

Abstract

Transposable elements (TEs) are major sources of genetic variation, and mating systems are believed to play a significant role in their dynamics. For example, insertion number is expected to be higher in sexual than in asexual organisms due to the inability of TEs to colonize new genomes in the absence of sex. The goal of this study was to determine the impact of the loss of sexual reproduction on TE load. Daphnia pulex has two reproductive modes, obligate and cyclical parthenogenesis, which differ with respect to the production of diapausing eggs. Cyclical parthenogens produce them meiotically, while obligate parthenogens produce them clonally. Pokey is a TTAA-specific DNA transposon, and is a stable component of Daphnia genomes. We used a PCR-based approach, TE-Display, to estimate the number of Pokey insertions in 22 cyclic and 22 obligate isolates of D. pulex. As expected, the copy number of Pokey insertions is significantly higher in cyclic than in obligate isolates. However, the distribution of elements among isolates within each breeding system is similar, which is congruent with the recent establishment of obligate lineages from a cyclic ancestor. We also assayed 46 isolates from eight cyclic populations and found that very few Pokey insertions were observed in more than one isolate, suggesting that Pokey has been active recently. Sequencing of PCR products from the TE-Display analysis shows that Pokey inserts into both coding and noncoding regions of the genome. However, there is no obvious similarity among sequences downstream of the TTAA Pokey insertion site.

Keywords

Transposable element Pokey Obligate parthenogenesis Cyclical parthenogenesis TE-Display 28S rRNA gene 

Introduction

Transposable elements (TEs) are a major source of genomic variation and their movement within and between genomes can have significant effects on their hosts if they insert into coding or regulatory regions of genes. The number of TEs in a species is a balance between natural selection, which typically decreases their copy number, and transposition, which increases their copy number. TEs are generally divided into two major classes defined by their sequence organization and mechanism of transposition (Finnegan 1990). Class I elements are typically called retrotransposons and move via an RNA intermediate, while Class II TEs, or transposons, move via a DNA intermediate.

Mating systems are believed to play a significant role in determining the fate of TEs (Wright and Schoen 1999). For example, they can increase their copy number within the genomes of asexual organisms but they are unable to move to new genomes (Hickey 1982). Conversely, TEs can spread through sexual populations by colonizing new genomes during zygote formation (Cavalier-Smith 1980). Thus, the number of TEs per individual is expected to be higher in organisms that produce sexually than asexually (Nuzhdin and Petrov 2003). Indeed, experimental studies in yeast have shown that TE frequency is lower in asexual populations than in sexual populations (Zeyl et al. 1996). Sullender and Crease (2001) found indirect evidence that the transposon, Pokey, transposes in Daphnia pulex that reproduce sexually, but not in those that lack sexual reproduction. Moreover, Arkhipova and Meselson (2005) found only one intact group of DNA elements, and Gladyshev et al. (2007) found only two types of intact LTR-retrotransposons in the ancient asexual Bdelloid rotifers. In addition, Wright et al. (2001) found a higher frequency of Ac-like insertions per individual in selfing populations of Arabidopsis thaliana than in outcrossing populations of the closely related Arabidopsis lyrata. The results of all these studies are consistent with the hypothesis that TE dynamics are likely to change with a change in reproductive mode.

Pokey is a Class II TE approximately 7 kbp in length that was originally detected in about 10% of the 28S ribosomal RNA (rRNA) genes of the cladoceran crustacean, Daphnia pulex (Sullender 1993). Pokey has 16-bp terminal inverted repeats and inserts into the sequence TTAA, which is duplicated upon insertion. Sequencing of two full-length Pokey elements from Daphnia pulicaria shows that it contains a single 1.5-kb open reading frame (ORF) that codes for a putative transposase (Penton et al. 2002). It is the only DNA transposon yet known to insert into 28S rRNA genes; other arthropod TEs that insert into these genes are non-LTR retrotransposons (Eickbush and Eickbush 2007). The conserved region of the 28S rRNA gene into which Pokey inserts is also a ‘‘hot spot’’ for insertional activity of these retrotransposons (Eickbush and Eickbush 2007). Since Pokey inserts into an rRNA gene, it is exposed to the same homogenization forces, such as gene conversion and unequal crossing-over, as are other rDNA-specific elements and the rRNA genes themselves (Eickbush and Eickbush 2007).

Pokey is a vertically inherited and stable element of Daphnia genomes (Penton and Crease 2004), which is unusual as DNA transposons usually require horizontal transfer to stay active (Kidwell and Lisch 2001). Moreover, Pokey also inserts into many other genomic locations (Sullender and Crease 2001), unlike the non-LTR retrotransposons that insert into 28S rRNA genes. To date, little is known about Pokey’s target site requirements outside of the 28S rRNA gene.

Daphnia are cladoceran crustaceans that inhabit freshwater lakes and ponds throughout the world. They typically reproduce by cyclical parthenogenesis in which direct-developing eggs are produced via apomictic parthenogenesis during favorable environmental conditions, and haploid resting eggs that require fertilization are produced when environmental conditions deteriorate. Direct-developing eggs can hatch into females or males, as sex determination is environmental. Some populations of Daphnia pulex reproduce by obligate parthenogenesis, in which the resting eggs are also produced via apomixis. Thus, these populations have lost the ability to reproduce sexually. The presence of both cyclic and obligate parthenogenesis in D. pulex makes it an ideal species with which to study the influence of breeding system on TE dynamics.

The goal of this study is to determine whether the loss of sexual reproduction in obligate parthenogens influences the number and position of Pokey elements in the genome relative to D. pulex that reproduce by cyclic parthenogenesis. In addition, we sampled multiple individuals from eight cyclical populations to determine the abundance and distribution of insertion sites within and between populations. In a previous study, Sullender and Crease (2001) used Southern blot analysis to characterize restriction fragment patterns of Pokey in D. pulex and found a significantly lower number of restriction fragments in obligate isolates than in cyclic isolates, as expected. In this study we use a PCR-based approach, TE-Display (described in more detail below), to detect Pokey elements in D. pulex from a much broader geographic range.

Methods

Daphnia pulex Samples and DNA Extraction

Daphnia pulex were sampled from ponds throughout Canada and the United States. In the lab, individual females were used to establish clonal lines of genetically identical offspring via apomixis. These samples are referred to as “isolates.” Animals were grown in artificial pond water and fed a mixture of one or both green algal species in the genera Chlamydomonas and Scenedesmus. Breeding system was diagnosed with a combination of allozyme analysis (Hebert et al. 1988, 1989; Hebert and Finston 2001) and examination of resting egg cases, which are readily produced in lab cultures when food becomes scarce. Cyclic females fail to release diapausing eggs into the egg case unless they have been fertilized, and thus the cases are empty when males are absent (Innes et al. 1986). Conversely, the egg cases of obligate females contain eggs even when males are absent. Some of the samples used in this study were generously provided by Michael Lynch (University of Indiana) and David Innes (Memorial University of Newfoundland), in whose labs the breeding system was diagnosed.

Daphnia pulex, which generally inhabit fishless ponds with substantial leaf litter, is known to hybridize with a closely related species, Daphnia pulicaria, which inhabits lakes and is able to coexist with fish. Such hybrids inhabit large shallow ponds with little leaf litter, and are generally obligate parthenogens (Hebert 1995). They can be identified using allozyme analysis of the enzyme lactate dehydrogenase (LDH; Hebert et al. 1993). All of the isolates used in this study were genotyped for LDH and those that were heterozygous for the species-specific alleles were not used in this study.

Genomic DNA was extracted from multiple individuals from each clonal line using one of three methods: phenol chloroform extraction (Crease 1986), CTAB extraction (Doyle and Doyle 1987), or extraction using a GenElute Mammalian Genomic DNA Miniprep kit (Sigma Aldrich, St. Louis, MO). The extractions were all done on about 10 to 30 fresh or frozen offspring from each culture, weighing approximately 100 mg (wet weight).

Eighty-three isolates from a range of geographical locations across North America (Fig. 1) were analyzed. Of these, 61 are cyclical parthenogens and they were sampled from 22 populations, as were the 22 obligate isolates. Thus, only one isolate was sampled from each of the obligate populations, and named O1 to O22. More than one isolate was sampled from 8 of the 22 cyclic populations, which are named C1 to C22. Isolates from the same population are named CX-Y, where X is the population number and Y is the isolate number (Table 1). For example, C4-2 is isolate 2 from cyclic population 4, while C4-5 is isolate 5 from the same population.
Fig. 1

General location of Daphnia pulex sampling sites. Squares indicate cyclically parthenogenetic populations and circles indicate obligately parthenogenetic populations. Numbers within symbols are the number of populations that were sampled from that province or state

Table 1

Number of Pokey insertions in 22 cyclically and 22 obligately parthenogenetic isolates of Daphnia pulex from eastern Canada and the United States (breeding system data set)

Isolate code

State/province

No. of Pokey insertions

C1

IL

9

C2-6

IL

6

C3-7

IL

10

C4-3

IL

7

C5-6

MI

4

C6-1

MI

3

C7

MI

5

C8-3

MI

3

C9

MI

4

C10-1

ON

7

C11-1

ON

1

C12

ON

5

C13

ON

3

C14

ON

3

C15

IN

5

C16

IN

4

C17

OR

5

C18

MN

5

C19

SK

5

C20

MB

9

C21

WI

4

C22

IA

7

O1

ME

2

O2

MI

3

O3

MI

7

O4

MI

5

O5

MI

2

O6

MI

8

O7

MI

4

O8

MI

1

O9

MN

2

O10

NB

3

O11

NS

2

O12

NY

1

O13

ON

1

O14

ON

2

O15

QC

3

O16

QC

8

O17

QC

4

O18

QC

2

O19

WI

3

O20

MI

3

O21

ON

2

O22

ON

4

Note: Each isolate represents a different population. When multiple isolates were analyzed from cyclically parthenogenetic populations, a single isolate was chosen randomly from each for inclusion in this data set. For example, C2–6 is one of seven isolates analyzed from population C2

The DNA samples were used to create two separate data sets: the “breeding system” and the “population genetics” data sets. The “breeding system” data set includes all 22 obligate parthenogens and one isolate from each of the 22 cyclic populations. A single isolate was chosen randomly from each of the eight cyclic populations from which multiple isolates were sampled (Table 1). The “population genetics” data set includes the 46 isolates from the eight cyclic populations from which multiple isolates were sampled (Table 2).
Table 2

The number of Pokey insertions in 46 isolates from eight cyclically parthenogenetic populations of Daphnia pulex from Illinois (IL), Michigan (MI), and Ontario (ON) (population genetics data set)

Isolate code

State/province

No. of Pokey insertions

C2-1

IL

1

C2-2

IL

1

C2-3

IL

3

C2-4

IL

6

C2-5

IL

1

C2-6

IL

6

C2-7

IL

1

C3-1

IL

4

C3-2

IL

1

C3-3

IL

2

C3-4

IL

7

C3-5

IL

12

C3-6

IL

9

C3-7

IL

10

C4-1

IL

8

C4-2

IL

6

C4-3

IL

7

C4-4

IL

1

C4-5

IL

9

C4-6

IL

5

C4-7

IL

4

C5-1

MI

10

C5-2

MI

3

C5-3

MI

6

C5-4

MI

4

C5-5

MI

3

C5-6

MI

4

C6-1

MI

3

C6-2

MI

4

C6-3

MI

1

C6-4

MI

1

C6-5

MI

11

C6-6

MI

9

C6-7

MI

7

C8-1

MI

5

C8-2

MI

6

C8-3

MI

3

C8-4

MI

4

C8-5

MI

7

C10-1

ON

7

C10-2

ON

8

C10-3

ON

2

C10-4

ON

6

C11-1

ON

1

C11-2

ON

1

C11-3

ON

4

Note: CX-Y is isolate Y from population X

Molecular Analyses

TE-Display of Pokey Insertion Sites

We used a PCR-based approach called TE-Display (TED) to examine the abundance and distribution of the transposon Pokey in the genome of D. pulex. This technique involves digestion of total genomic DNA with a frequent-cutting restriction enzyme, followed by the ligation of adaptors to the digested fragments. This digestion-ligation template is then PCR amplified using a TE-specific primer and an adaptor-specific primer to generate fragments containing genomic sequences that are immediately downstream of TE insertions (Wright et al. 2001).

We digested approximately 100 ng of genomic DNA with a 260/280 ratio of between 1.6 and 2 with 5 units of BfaI restriction enzyme (New England Biolabs, Ipswich, MA) in a total volume of 50 μl for 5 h at 37°C in 1× reaction buffer supplied by the manufacturer. If the 260/280 ratio of a genomic DNA sample was ≥2, indicating the presence of RNA, we digested about 400 ng of the sample. The digested DNA was ethanol precipitated, resuspended in 15 μl of sterile water, and then ligated to 25 pmol of forward and reverse BfaI linkers (Table 3) in a 20-μl reaction containing 400 units of T4 DNA Ligase (New England Biolabs, Ipswich, MA) in 1× T4 DNA Ligase buffer. The ligation reaction was incubated overnight at 4°C.
Table 3

Sequences of primers and linkers used in this study

Primer

Sequence

Purpose

Pok6172F

5′-tggtcgatggtaaagacctcaacgtc

PCR with Pok6578R

Pok6578R

5′-gatggtcggattcgattgaatgctcg

PCR with Pok6172F

Pok6456F

5′-gacaacggtggccgaaacgcgg

Primary PCR for TED with BfaIR

HEX-Pok6464F

5′-tggccaaaacacggtttggccg

Secondary PCR for TED with BfaIR

Pok6466F

5′-gccgaaacgcggttaggccg

Sequencing of plasmid clones

28S3257R

5′-cgtctcccacttatgctacacctc

PCR with Pok6466F

BfaIR

5′-gacgatgagtcctgagtag

Primary and secondary PCR for TED

BfaI linkerF

5′-tactcaggactcat

TED

BfaI linkerR

5′-gacgatgagtcctgag

TED

Note: Primers that anneal to the transposon, Pokey (Pok), are numbered according to their nucleotide position in a full-length element from Daphnia pulicaria (accession no. AY115589 [Penton et al. 2002]). Primers that anneal to the 28S rRNA gene (28S) are numbered according to their position in the 28S rRNA gene of Daphnia pulicaria (accession no. AF346514 [Omilian and Taylor 2001]). All oligonucleotides were purchased from Sigma Aldrich (St. Louis, MO). TED, Transposable Element-Display

The ligated DNA was amplified in a primary (preselective) PCR reaction of 50 μl total volume that contained 1 μl of the digested-ligated DNA sample, 1 pmol of the primer Pok6456F (located near the 3′ end of the Pokey element), 1 pmol of the primer BfaIR (Table 3), 2 mM MgCl2, 1× PCR buffer (10 mM TrisHCl, pH 8.3, 50 mM KCl), 0.8 mM dNTPs, and 0.5 unit of Taq DNA polymerase (New England Biolabs). PCR reactions were performed in a PTC-100 Thermocycler (MJ Research, Waltham, MA) for 35 cycles each consisting of 30 s of denaturing at 94°C, 90 s of annealing at 55°C, and 1 min of extension at 72°C, with a final extension of 5 min at 72°C.

We determined the DNA concentration of the primary PCR products using a Nanodrop ND-1000 spectrophotometer. The concentration was consistently found to be 120–140 ng/μl. Approximately 60–70 ng of the primary PCR product was used as the template for a secondary (selective) PCR amplification. This reaction was carried out under the same reaction conditions as above except that the total volume was 20 μl, only 0.25 unit of Taq DNA polymerase (New England Biolabs) was used, and the forward primer was Pok6464F (Table 3) labeled with the fluorescent dye, HEX (Applied Biosystems, Foster City, CA). This primer is just downstream of Pok6456F, at the 3′ end of the Pokey element.

The secondary PCR products were resolved on an ABI 3730 DNA Analyzer (Applied Biosystems, Foster City, CA) in the Genomics Facility at the University of Guelph. Two microliters of each secondary PCR product was loaded into a 96-well plate. A size standard was prepared by mixing 4 μl of ROX-labeled sizing ladder (Applied Biosystems) containing fragments ranging from 35 to 500 bp with 750 μl of formamide and then adding 15 μl to each sample in the 96-well plate. The samples were then denatured at 96°C for 5 min and placed on ice for 5 min before loading them into the ABI 3730. The GeneMapper 50_POP7_template_1 run module was used to analyze the samples.

GeneMapper Software v3.7 (Applied Biosystems, Foster City, CA) was used to analyze the electropheretograms generated by the ABI 3730. The software estimates the size of the peaks in the electropheretogram by comparing their mobility to the known sizes of the fragments in the ROX sizing ladder and then generates a list of fragments and their sizes for each sample. We only included fragments with a peak height of 100 fluorescence units in subsequent analyses.

To ensure that the TED fragment patterns are reproducible, we repeated each individual primary PCR reaction three times for each isolate. A separate secondary PCR reaction was performed from each of the triplicate primary PCR reactions, and the triplicate electropheretograms for each sample were compared to each other. If at least two of the triplicate primary PCR reactions showed the same fragment pattern, then they were included in the analyses. If there were not at least two similar electropheretograms for any sample, we repeated the TED with new triplicates. This situation occasionally occurred when the amount of genomic DNA in the primary PCR reactions was not sufficient, resulting in weak peaks in two of the three replicates (<100 fluorescence units). In these cases, new triplicates were generated using 50% more genomic DNA template in the primary PCRs.

If there is a BfaI restriction site immediately downstream of the TTAA Pokey insertion site, a fragment of 160 bp should be generated by TED. Thus, although there were some fragments smaller than 160 bp, we only included fragments ≥160 bp in the analysis. The largest fragment observed was 528 bp. We used the results to generate a Pokey profile for each isolate indicating the presence (1) or absence (0) of all the insertion fragments observed (Supplementary Table 1).

Cloning and Sequencing of TED Fragments

In order to confirm that the PCR fragments generated by TED are truly Pokey insertions, and to determine if fragments of very similar size in different isolates are the same Pokey insertion site, we cloned the secondary PCR products of six isolates (C4-1, C5-1, C6-7, C17, C20, O3). We chose these isolates because they represented much of the Pokey insertion site variation observed among isolates. The fragments were cloned using the StrataClone Blunt PCR Cloning Kit (Stratagene, La Jolla, CA) following the manufacturer’s specifications. We chose 48 colonies from each experiment and grew them overnight at 37°C in 1.5 ml of Luria Broth with ampicillin (200 mg/L). Plasmid DNA was isolated from these cultures using the Montage Plasmid Miniprep kit (Millipore, Billerica, MA) according to the manufacturer’s specifications. The final pellet was dissolved in 50 μl of water.

The size of the plasmid inserts was estimated by amplifying them with the M13F and M13R primers (Stratagene) and then resolving them on a 0.8% agarose gel in 1× TEA buffer. The gel was stained with ethidium bromide, and the PCR products were visualized under UV light. We chose at least one plasmid clone of each size from each isolate, as well as clones of similar size from different isolates, to determine if they contain the same Pokey insertion sites. Eight microliters of plasmid DNA was sequenced in a reaction containing 10 pmol of the Pok6466F primer (Table 3), 0.5 μl of Big Dye Terminator sequencing mix (Applied Biosystems), and 1 μl of Big Dye sequencing buffer (Applied Biosystems) in a total volume of 12 μl. The sequencing reaction was performed on a PTC-100 Thermocycler (MJ Research, Waltham, MA) for 55 cycles, each consisting of 30 s of denaturing at 96°C, 15 s of annealing at 55°C, and 4 min of extension at 60°C. The samples were then resolved on an ABI 3730 DNA Analyzer (Applied Biosystems) at the Genomics Facility at the University of Guelph. The sequences were compared to the D. pulex genome sequence, available at http://www.genome.jgi-psf.org/Dappu1/Dappu1.home.html, using the BLASTN procedure (Altschul et al. 1997) to determine what the genomic flanking regions might be.

PCR Amplification of Pokey from Genomic DNA

To confirm the insertion of Pokey elements in 28S rRNA genes, we PCR-amplified genomic DNA from some D. pulex isolates using primers Pok6466F and 28S3257R to generate an ∼340-bp fragment containing the 3′ end of Pokey and the downstream flanking region of the 28S rRNA gene. Amplification using primers Pok6172F and Pok6578R, which generates an ∼400-bp internal Pokey fragment, was also performed on the same isolates as a positive control. The amplification reactions were performed on a PTC-100 Thermocycler (MJ Research) for 35 cycles, each consisting of 30 s of denaturing at 94°C, 90 s of annealing at 55°C, and 1 min of extension at 72°C, and a final extension of 5 min at 72°C. The PCR products were resolved on 0.8% agarose gels in 1× TEA, stained with ethidium bromide, and photographed under UV light.

Population Genetics Analysis

After obtaining the Pokey insertion fragment profile for each of the Daphnia isolates, we used the software package, Arlequin version 3.11 (Excoffier et al. 2005) to partition the variation in Pokey insertion sites within and among isolates from different breeding systems and among populations of cyclic parthenogens. Genetic distance was calculated between all pairs of Pokey profiles as dxy = Σδxy(i), where δxy(i) is equal to 0 if the alleles of the ith locus are identical for both profiles, and equal to 1 if they are different (Excoffier et al. 2005). The resulting distance matrix was used to draw a neighbor joining (NJ) tree using MEGA v3.0 software (Kumar et al. 2004). A t-test was used to determine if the difference in mean number of Pokey insertions differed between obligately and cyclically parthenogenetic isolates. We also used Arlequin to perform Tajima’s test of neutrality (D statistic) for the cyclic populations.

Results

Cloning of TED Fragments

In order to ensure that the TED technique identifies Pokey insertions, we cloned and then sequenced the secondary PCR products from six isolates with different Pokey profiles. These sequences only contain the sequence downstream of the Pokey insertion sites. Thirty-eight clones were chosen from the six experiments for sequencing. Three of the clones were excluded from further analysis as their sequences revealed that they did not contain an insert. The other 35 sequenced fragments are indeed Pokey insertions and range in size between 163 and 704 bp, indicating that there are some TED fragments larger than those we could reliably score using our ROX size standard. Fragments of the same size in the electropherograms generated by the ABI 3730 DNA analyzer do represent the same Pokey insertion site (Supplementary Table 2). For example, the Pokey profile of isolates C6-7 and O3, which are from two different breeding systems, both include a fragment of 261 bp and their sequences are the same. The same situation was observed in C6-7 and C5-1, which are two isolates from different populations, but the same breeding system. There were cases in which two TED fragments differ in size by only a couple of nucleotides and they are in fact different insertions (Supplementary Table 2).

The results of this analysis also indicate that TTAA is not the only site into which Pokey is inserted. Nine of the cloned flanking sequences (three from O3, two from C4-1, three from C6-7, and one from C17) are the 28S rRNA gene with the expected TTAA insertion site (Sullender and Crease 2001). Although most of the other clones also have a TTAA insertion site, six have ATAA, and five have TTAT. We also observed one each of TGAA, TTGG, TTAC, and TTAG (Supplementary Table 2). Unexpectedly, Pokey is inserted into the 28S rRNA gene in the opposite orientation in one clone from isolate C4-1 (Supplementary Table 2).

All but 4 of the other 25 flanking sequences were compared to the D. pulex genome sequence (Supplementary Table 2). These four are very short (38, 13, or 9 nt) and were therefore not analyzed. Of the others, six, including one from the obligate isolate (O3), show significant matches to putative coding regions. The other sequences match introns or other noncoding regions. One sequence contains the 5S rRNA gene, although the element is not located in the gene itself (Supplementary Table 2). Three clones show similarity to a D. pulex microsatellite sequence (accession no. AY619285); two share the same flanking sequence, but the third is different from the other two.

Breeding System Data Set

The breeding system data set contains 22 cyclically and 22 obligately parthenogenetic isolates, each from a different population. The mean number of Pokey insertions in cyclic isolates is 5.18 (±2.24), with values from 1 to 10, while the mean number of Pokey insertions in obligate isolates is only 3.27 (±2.07), with values from 1 to 8. The mean number of insertions in cyclic isolates is significantly higher, as expected (t-statistic = 2.93, p [one-tail] = 0.002).

Overall, 59 different Pokey insertion sites were detected. The 22 obligate isolates show 31 of these Pokey insertion sites, of which 17 (55%) are found in only one isolate (Fig. 2). The 22 cyclic isolates display 48 of these insertion sites, of which 28 (58%) are found in only one isolate. Twenty of the 59 insertion sites (34%) were observed in at least one obligate and one cyclic isolate. This includes the 242-bp fragment generated by insertions in 28S rRNA genes, which are present in 37 of 44 isolates. However, more than half of the insertion sites are present in only one or two isolates in each breeding system (Fig. 2).
Fig. 2

Frequency distribution of Pokey insertions in 22 cyclically (black) and 22 obligately (gray) parthenogenetic isolates of Daphnia pulex from eastern Canada and the United States (breeding system data set)

Five isolates without Pokey insertions in 28S rRNA genes (denoted 242-0) and three isolates with insertions in 28S rRNA genes (denoted 242-1) were amplified with two pairs of primers. Pok6172F and Pok6578R generate a ∼400 bp internal Pokey fragment while Pok6466F and 28S3257R generate an ∼340-bp fragment containing the 3′ end of Pokey and the 28S rRNA gene downstream of the element. As expected, the three 242-1 isolates all amplified strongly with both primer pairs (data not shown). Three of the five 242-0 isolates also amplified with the Pok/28S primer pair, although not as strongly as the 242-1 controls. One 28S rRNA gene fragment from each of a 242-1 and a 242-0 isolate was sequenced in both directions, which confirmed that the expected fragment was amplified and that there is no difference in the location of BfaI restriction sites in the two isolates (data not shown).

The AMOVA results for the breeding system data set show that 96.4% of the variation in Pokey insertion occurs within groups (cyclic and obligate), with little additional variation among groups, although it is significant (FST = 0.036, < 0.001; Table 4). There are only a few small clusters of cyclic isolates in the NJ tree (Fig. 3), but otherwise we did not observe substantial clustering of Pokey insertion profiles from the same breeding system or geographical location.
Table 4

AMOVA of Pokey insertion site variation in 22 cyclically and 22 obligately parthenogenetic isolates of Daphnia pulex from eastern Canada and the United States (breeding system data set)

Source of variation

df

Sum of squares

Variance component

Percentage variation

p value

Among groups

1

5.77

0.12

3.63

0.00098

Within groups

42

132.63

3.16

96.37

 

  Total

43

138.40

3.28

  

  FST

0.036

   

0.00098

Note: Significance tests were based on 1023 permuations. df, degrees of freedom

Fig. 3

Unrooted neighbor-joining tree of Pokey insertion profiles from 22 cyclically parthenogenetic (C) and 22 obligately parthenogenetic (O) isolates of Daphniapulex from eastern Canada and the United States (breeding system data set)

Population Genetics Data Set

The population genetics data set contains 46 cyclic isolates from eight different populations, divided into three groups by geographic origin: Illinois, Michigan, and Ontario (Table 2). AMOVA for these three groups shows that most of the variation in Pokey insertion sites occurs within populations (89.3%), with little additional variation among isolates within groups or among groups, although significant variation among populations within groups (6%) was observed (Table 5). Variance among groups (5%) is not significant (p = 0.06). Moreover, Pokey profiles from the same population or from populations that are geographically close to one another do not tend to cluster together on the NJ tree (Fig. 4).
Table 5

AMOVA of Pokey insertion site variation in eight cyclically parthenogenetic populations of Daphnia pulex from Illinois, Michigan, and Ontario (population genetics data set)

 

df

Sum of squares

Variance component

Percentage variation

p value

Source of variation

  Among groups

2

13.19

0.16 Va

4.75

0.06

  Among populations within groups

5

21.80

0.21 Vb

5.96

0.01

  Within population

38

119.10

3.13 Vc

89.30

<0.0001

  Total

45

154.09

3.50981

 

Fixation indexes

  FSC

0.062

Among populations within groups

0.01

  

  FCT

0.047

Among groups

0.06

  

Note: Significance tests were based on 1023 permuations. df, degrees of freedom

Fig. 4

Unrooted neighbor-joining tree of Pokey insertion profiles of Daphniapulex sampled from eight cyclically parthenogenetic populations in Illinois, Michigan, and Ontario (population genetics data set)

The 46 cyclic isolates in the population genetics data set have 58 different Pokey insertion sites. Twenty-five sites (43%) were observed in more than one population, but most sites (70%) were found in only one or two isolates (Fig. 5). The highest number of isolates is 40, which is the 28S rRNA gene (242-bp) insertion. Other sites are present in most populations, but in fewer isolates (Supplementary Table 1). For example, sites 201 and 260 bp were observed in six of the eight populations (12 isolates), site 192 bp was observed in seven populations (18 isolates), and site 308 bp was observed in all eight populations (23 isolates).
Fig. 5

Frequency distribution of 58 Pokey insertions in 46 cyclically parthenogenetic isolates of Daphnia pulex from eight populations in Illinois, Michigan, and Ontario (population genetics data set)

Estimates of Tajima’s D are negative in six populations, positive in one population, and equal to zero in one population. The negative D values are consistent with the operation of purifying selection, but none of the values are significant (data not shown), which may be partly due to the small sample sizes. However, based on these results, we cannot reject the hypothesis that the variation in Pokey site insertion is neutral even though the overall pattern we observed suggests that purifying selection is operating.

Discussion

The lack of recombination between genomes in asexual organisms is the basis of the hypothesis that TEs should behave differently in sexual and asexual breeding systems (Hickey 1982). The results of this study support the hypothesis that breeding system will impact Pokey dynamics in D. pulex. On average, the number of Pokey insertions is significantly higher in cyclic than in obligate parthenogens, as predicted. Insertion site variation among isolates is similar in both breeding systems, but only 27% of sites were observed in both cyclic and obligate parthenogens. These findings do not support the hypothesis that lack of transposition in obligate parthenogens will result in lower levels of Pokey insertion site variation among isolates. However, they do suggest that Pokey does not remain in one location for long periods of time, either because Pokey insertions are quickly lost by natural selection or genetic drift or because Pokey is very active.

The hypothesis that Pokey is active is consistent with the study by Wright et al. (2001) in which they used TED to determine the population dynamics of a DNA transposon, Ac-III, in self-pollinating Arabidopsis thaliana and its close outcrossing relative, Arabidopsis lyrata. They provided evidence that Ac-III has recently been active in Arabidopsis and observed a frequency distribution of the element in outcrossing populations (their Fig. 3a) that is very similar to the one we observed for Pokey (Figs. 2 and 5). This is consistent with the suggestion by Sullender and Crease (2001) that Pokey is active in cyclic parthenogens. However, this also raises the question of why each obligate isolate has so many unique Pokey insertion sites if the element is not active in obligate parthenogens (Sullender and Crease 2001). One explanation is that obligately parthenogenetic populations are very recently derived from cyclical parthenogens, and thus the former carry unique insertions that were recently inherited from the latter. Thus, the overall Pokey insertion profile is similar in obligate and cyclic isolates because each obligate clone has “captured” a sample of Pokey insertions from the cyclic population from which it was derived. This is consistent with the results of Crease et al. (1989), who analyzed population structure of D. pulex from Ontario and Michigan using analysis of sequence variation in mitochondrial genes. They concluded that obligate parthenogenesis has a polyphyletic origin in D. pulex and that obligate clones are evolutionarily young. Paland et al. (2005) did a more extensive survey of D. pulex populations in eastern Canada and the United States and, also, concluded that obligate lineages of D. pulex are polyphyletic. Moreover, they estimated the mean maximum age of clonal lineages to be less than 200,000 years.

Given the polyphyletic and recent origin of obligate clones from cyclic populations, the lower level of Pokey insertion in obligate parthenogens could be explained by interclonal selection after the loss of sex, such that clonal lines that inherit lower insertion loads from their cyclic progenitors are more likely to persist than those that inherit higher loads (Arkhipova and Meselson 2004). It is also possible that Pokey insertions could be lost within clonal lines after the loss of sex as a result of interchromosomal recombination during apomixis, which is known to occur in D. pulex (Omilian et al. 2006).

An extreme example of interclonal selection occurs in the Bdelloid rotifers, which are likely to represent a truly ancient sexual lineage. Arkhipova and Meselson (2000) were unable to detect active retrotransposon families in these organisms despite the occurrence of many groups in rotifers that reproduce sexually. More recently, Gladyshev et al. (2007) described two types of intact LTR-retrotransposons in Bdelloids that may be active, but have likely been recently acquired. Moreover, the only putatively active DNA transposons that have been found in these organisms are mariner elements that were also likely to have been recently acquired, possibly by horizontal transfer (Arkhipova and Meselson 2005). The hypothesis of interclonal selection could be tested in D. pulex by measuring TE load in a large number of obligate clones of different ages. For example, Paland et al. (2005) showed that the oldest obligate lineages occur in the Maritime provinces and the northeastern United States, with clone age decreasing as one moves southwest toward the Great Lakes region and into the Midwestern United States. Thus, lineages sampled from the Northeast would be expected to have lower Pokey loads than those from the Midwest.

The NJ tree of Pokey profiles from cyclic and obligate isolates (Fig. 3) shows that, in general, there is little clustering of profiles within each of the two breeding systems. This result does not support the suggestion that there is a more conserved Pokey insertion pattern within each breeding system and is consistent with the polyphyletic origin of obligately parthenogenetic clones from cyclical parthenogens.

Similarity among Pokey insertion profiles is not associated with the geographical distance between populations since shared insertion sites are found in geographically distant populations. All but one insertion site that was found in 5 or more cyclic isolates in the breeding system data set was also found in 10 or more isolates and more than one population in the population genetics data set (Supplementary Table 1). Moreover, four sites were found in six or more of the eight cyclic populations. One explanation for these results is that some copies of Pokey persist in the same genomic location for substantial periods of time, even though other copies are active. Alternatively, these locations may be hotspots for Pokey insertion, as are the 28S rRNA genes. Thus, some of the elements in these sites could represent independent insertion events.

Sullender (1993) previously reported that Pokey is flanked by a TTAA site, but in the current study, Pokey insertions were also found in other sites. There are two explanations for this: either TTAA is not the only site into which Pokey can insert or Pokey originally inserted into a TTAA site but the site was subsequently mutated. A preliminary search of the D. pulex genome sequence (T. Crease, unpublished data) identified seven full-length Pokey elements. Three are flanked by TTAA at both ends, as expected, while one is flanked by TTAA and ATAA and one is flanked by CTAA and ATAA, suggesting that mutation may have altered the insertion site at one or both ends. However, two of these elements have CTAA at both ends, which suggests that Pokey may be able to insert at sites other than TTAA.

We are not aware of any data suggesting that Pokey can insert in the reverse direction in 28S rRNA genes, but we identified one such insertion among our cloned TED fragments. We also observed some isolates from cyclic populations that showed no 242-bp fragment (242-0) in the TED analysis even though a Pok/28S fragment was amplified using the 28S3257R primer. A change in the BfaI site downstream of Pokey’s insertion site could have explained why we did not observe a 242-bp fragment in TED even though Pokey is present in some copies of the 28S rRNA gene, although the cloned fragment that we sequenced did possess a BfaI site in the expected location. Moreover, there is no association between the amplification of the Pok/28S fragment and the number of genomic elements detected in 242-0 isolates. As expected, the two obligate isolates we tested did not amplify with the Pok/28S primers. One of these had three and the other had two additional TED fragments. Conversely, the three cyclic 242-0 isolates we tested all amplified with the Pok/28S primers and had 10, 6, and 1 additional TED fragments. Indeed, the strongest amplification of the Pok/28S fragment occurred in the isolate with 10 additional TED fragments. Thus, it is not clear why TED failed to detect 28S rRNA gene insertions in these isolates.

Overall, we observed three obligate isolates that did not contain the 242-bp fragment, two of which were tested with Pok/28S primers and failed to amplify, as expected. In addition, all of these isolates had a small number of additional insertions. Given that the majority of cyclic populations have an abundance of Pokey insertions, and at least some of the elements appear to be active, it is unlikely that Pokey would be lost completely from a cyclic population. However, these results do suggest that obligate clones completely free of Pokey insertions, even in 28S rRNA genes, may exist. In fact, if Pokey is not active in obligate parthenogens as suggested by Sullender and Crease (2001), one would expect most obligate clones to be free of Pokey insertions in 28S rRNA genes because concerted evolution of rDNA continues to occur in clonal lineages despite the lack of meiosis (Crease and Lynch 1991). Indeed, substantial changes in the relative frequency of rDNA variants can occur in less than 20 generations of apomictic reproduction (McTaggart et al. 2007). As suggested above, these results suggest that the obligate clones we analyzed are very recently derived from a cyclic ancestor and/or that Pokey is active even in obligate parthenogens.

Our BLAST search of the D. pulex genome with cloned TED fragments showed that Pokey does insert into coding sequences, so these insertions are potentially deleterious. Thus, the persistence of obligate clones is expected to be associated with a low level of insertion into coding regions, in addition to a low insertion load overall. A more thorough comparison of Pokey insertion site variation in cyclic populations and in obligate clones of different ages would test this hypothesis.

Overall, our results are consistent with studies of yeast (Zeyl et al. 1996) and rotifers (Arkhipova and Meselson 2000, 2005), as well as results from earlier work on D. pulex (Sullender and Crease 2001), which all suggest that TE copy number is higher in organisms that reproduce sexually compared to their asexual relatives. However, in the yeast study (Zeyl et al. 1996), as well as the assessment of Ac-III in outbred and inbred Arabidopsis (Wright et al. 2001), the number of TE insertions was compared in two different species with different breeding systems. Thus, factors other than breeding system may have caused the variable abundance of TEs in the two species. The present study was performed on isolates from the same species but two different breeding systems, as was the study by Sullender and Crease (2001). The difference between our assessment and theirs is the method that we used to detect Pokey insertions. They used Southern blot analysis to characterize restriction fragment patterns of Pokey and found an average of 16.6 restriction fragments in cyclic isolates and 15.7 fragments in obligate isolates. In contrast, we found an average of 5.2 fragments in cyclic isolates and 3.3 fragments in obligate isolates using TED. We scored only fragments that are 500 bp or smaller due to the limitation of the size standard that was available, and our cloning results suggest that we did fail to detect at least some larger fragments in the TED analysis, but this would likely have little impact on our conclusions. Moreover, TED only detects Pokey insertions that match the primer sites at the 3′ end of the element, so it will not detect elements with degraded 3′ ends. In contrast, a 1.1-kbp probe, which could bind to degraded elements even if there are mismatches, was used in Sullender and Crease’s (2001) Southern blot analysis. Consequently, this approach could detect both active and inactive or degraded Pokey insertions, while TED is more likely to detect only active (or recently inactivated) Pokey insertions. Thus, an advantage of TED is its failure to detect highly degraded elements, which are least likely to affect the host. This idea is supported by a preliminary search for Pokey elements in the D. pulex genome sequence (T. Crease, unpublished data), which identified fewer than 10 intact elements, but many partial elements that are missing one or both of the 5′ and 3′ terminal inverted repeats that are required for transposition (Finnegan 1990).

Overall, this study suggests that breeding system does affect the dynamics of TEs, as there are more Pokey insertion sites in cyclically parthenogenetic than in obligately parthenogenetic D. pulex. In addition, it appears that Pokey is active and often moves to new locations so that patterns of insertion among isolates are not correlated with the breeding system or the geographical location of the host populations. This finding is congruent with the work of Crease et al. (1989) and Paland et al. (2005) suggesting that obligately parthenogenetic populations of D. pulex were very recently derived from cyclical parthenogens. However, it is not congruent with the suggestion that Pokey is inactive in obligate parthenogens (Sullender and Crease 2001). Future work will be aimed at measuring the transposition rate in laboratory cultures during both obligate and cyclic parthenogenesis.

Notes

Acknowledgments

Financial support for the work was provided by a research grant from the Natural Sciences and Engineering Research Council of Canada to T.J.C. We thank M. Lynch and D. Innes for generously providing some of the D. pulex isolates used in this study and A. Holliss and E. Holmes for analysis of TED fragments and sequencing of plasmid clones on the ABI3730. Comments provided by two anonymous reviewers improved an earlier version of the paper.

Supplementary material

239_2008_9118_MOESM1_ESM.xls (162 kb)
(XLS 162 kb)
239_2008_9118_MOESM2_ESM.doc (52 kb)
(DOC 52 kb)

References

  1. Altschul S, Madden T, Schaffer A, Zhang J, Zhang Z, Miller W, Lipman D (1997) Gapped BLAST and PSI-BLAST: a new generation of protein data base search program. Nucleic Acids Res 25:3389–3402PubMedCrossRefGoogle Scholar
  2. Arkhipova I, Meselson M (2000) Transposable elements in sexual and ancient asexual taxa. Proc Natl Acad Sci USA 97:14473–14477PubMedCrossRefGoogle Scholar
  3. Arkhipova I, Meselson M (2004) Deleterious transposable elements and the extinction of asexuals. Bioessays 27:76–85CrossRefGoogle Scholar
  4. Arkhipova I, Meselson M (2005) Diverse DNA transposons in rotifers of the class Bdelloidea. Proc Natl Acad Sci USA 102:11781–11786PubMedCrossRefGoogle Scholar
  5. Cavalier-Smith T (1980) How selfish is DNA? Nature 285:617–618PubMedCrossRefGoogle Scholar
  6. Crease T (1986) Mitochondrial DNA variation in Daphnia pulex Leydig producing by obligate and cyclic parthenogenesis. Ph.D. dissertation, Washington University, St. Louis, MOGoogle Scholar
  7. Crease T, Lynch M (1991) Ribosomal DNA variation in Daphnia pulex. Mol Biol Evol 8:620–640Google Scholar
  8. Crease T, Stanton D, Hebert P (1989) Polyphyletic origin of asexuality in Daphnia pulex. II. Mitochondrial DNA variation. Evolution 43:1016–1026CrossRefGoogle Scholar
  9. Doyle J, Doyle J (1987) A rapid DNA isolation procedure for small quantities of fresh leaf. Phytochem Bull 19:11–15Google Scholar
  10. Eickbush T, Eickbush D (2007) Finely orchestrated movements: evolution of the ribosomal RNA genes. Genetics 175:477–485PubMedCrossRefGoogle Scholar
  11. Excoffier L, Laval G, Schneider S (2005) Arlequin ver. 3.0: an integrated software package for population genetics data analysis. Evol Bioinform Online 1:47–50Google Scholar
  12. Finnegan D (1990) Transposable elements and DNA transposition in eukaryotes. Curr Opin Cell Biol 2:471–477PubMedCrossRefGoogle Scholar
  13. Gladyshev E, Meselson M, Arkhipova I (2007) A deep-branching clade of retrovirus-like retrotransposons in Bdelloid rotifers. Gene 390:136–145PubMedCrossRefGoogle Scholar
  14. Hebert P (1995) The Daphnia of North America. CD-ROM. University of Guelph, Guelph, ONGoogle Scholar
  15. Hebert P, Beaton M, Schwartz S, Stanton D (1989) Polyphyletic origins of asexuality in Daphnia pulex: I. Breeding system variation and levels of clonal diversity. Evolution 43:1004–1015CrossRefGoogle Scholar
  16. Hebert P, Finston T (2001) Macrogeographic patterns of breeding system diversity in the Daphnia pulex group from the United States and Mexico. Heredity 87:153–161PubMedCrossRefGoogle Scholar
  17. Hebert P, Ward R, Weider L (1988) Clonal diversity patterns and breeding system variation in Daphnia pulex: an asexual–sexual complex. Evolution 42:147–159CrossRefGoogle Scholar
  18. Hebert P, Schwartz S, Ward R, Finston T (1993) Macrogeographic patterns of breeding system diversity in the Daphnia pulex group. I. Breeding systems of Canadian populations. Heredity 70:148–161CrossRefGoogle Scholar
  19. Hickey D (1982) Selfish DNA: a sexually transmitted nuclear parasite. Genetics 101:519–531PubMedGoogle Scholar
  20. Innes D, Schwartz S, Hebert P (1986) Genotypic diversity and variation in mode of reproduction among populations in the Daphnia pulex group. Heredity 57:345–355CrossRefGoogle Scholar
  21. Kidwell M, Lisch D (2001) Perspective: transposable elements, parasitic DNA, and genome evolution. Evolution 55:1–24PubMedGoogle Scholar
  22. Kumar S, Tamura K, Nei M (2004) MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief Bioinform 5:150–164PubMedCrossRefGoogle Scholar
  23. McTaggart S, Dudycha J, Omilian A, Crease T (2007) Rates of recombination in the ribosomal DNA of apomictically propagated Daphnia obtusa lines. Genetics 175:311–320PubMedCrossRefGoogle Scholar
  24. Nuzhdin S, Petrov D (2003) Transposable elements in clonal lineages: lLethal hangover from sex. Biol J Linn Soc 79:33–41CrossRefGoogle Scholar
  25. Omilian A, Taylor D (2001) Rate acceleration and long-branch attraction in a conserved gene of cryptic daphniid (Crustacea) species. Mol Biol Evol 18:2201–2212PubMedGoogle Scholar
  26. Omilian A, Cristescu M, Dudycha J, Lynch M (2006) Ameiotic recombination in asexual lineages of Daphnia. Proc Natl Acad Sci USA 103:18638–18643PubMedCrossRefGoogle Scholar
  27. Paland S, Colbourne J, Lynch M (2005) Evolutionary history of contagious asexuality in Daphnia pulex. Evolution 59:800–813PubMedGoogle Scholar
  28. Penton E, Crease T (2004) Evolution of the transposable element Pokey in the ribosomal DNA of the species in the subgenus Daphnia (Crustacea: Cladocera). Mol Biol Evol 21:1727–1739PubMedCrossRefGoogle Scholar
  29. Penton E, Sullender B, Crease T (2002) Pokey, a new DNA transposon in Daphnia (Cladocera: Crustacea). J Mol Evol 55:664–673PubMedCrossRefGoogle Scholar
  30. Sullender B (1993) Preliminary characterization and population genetic survey of the Daphnia rDNA transposable element, Pokey. Ph.D. dissertation, University of Oregon, Eugene, Eugene, ORGoogle Scholar
  31. Sullender B, Crease T (2001) The behavior of a Daphnia pulex transposable element in cyclically and obligately parthenogenetic populations. J Mol Evol 53:63–69PubMedGoogle Scholar
  32. Wright S, Schoen D (1999) Transposon dynamics and the breeding system. Genetics 107:139–148CrossRefGoogle Scholar
  33. Wright S, Le QH, Schoen D, Bureau T (2001) Population dynamics of an Ac-like transposable element in self- and cross-pollinating Arabidopsis. Genetics 158:1279–1288PubMedGoogle Scholar
  34. Zeyl C, Bell G, Green D (1996) Sex and the spread of retrotransposon Ty3 in experimental populations of Saccharomyces cerevisiae. Genetics 143:1567–1577PubMedGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2008

Authors and Affiliations

  1. 1.Department of Integrative BiologyUniversity of GuelphGuelphCanada

Personalised recommendations