Report

Trans-kingdom Transposition of the Drosophila Element mariner Within the Protozoan Leishmania

See allHide authors and affiliations

Science  13 Jun 1997:
Vol. 276, Issue 5319, pp. 1716-1719
DOI: 10.1126/science.276.5319.1716

Abstract

Transposable elements of the mariner/Tc1 family are postulated to have spread by horizontal transfer and be relatively independent of host-specific factors. This was tested by introducing the Drosophila mauritiana element mariner into the human parasite Leishmania major, a trypanosomatid protozoan belonging to one of the most ancient eukaryotic lineages. Transposition in Leishmania was efficient, occurring in more than 20 percent of random transfectants, and proceeded by the same mechanism as in Drosophila. Insertional inactivation of a specific gene was obtained, and a modified mariner element was used to select for gene fusions, establishing mariner as a powerful genetic tool for Leishmania and other organisms. These experiments demonstrate the evolutionary range ofmariner transposition in vivo and underscore the ability of this ubiquitous DNA to parasitize the eukaryotic genome.

Transposons of themariner/Tc1 family are ubiquitous elements of eukaryotic genomes, occurring in virtually every taxon examined (1-3). Phylogenetic studies of mariner elements have provided compelling evidence for the occurrence of horizontal transfer across species during evolution, traversing distances as far as that separating insects and flatworms (1, 2, 4). This suggested that mariner could transpose independently of host-specific factors, a belief bolstered by studies of transposition activity in vitro (5). Hence, mariner was advanced as a potentially general tool for stable transformation and insertional mutagenesis in eukaryotic genomes after heterologous expression (2, 3, 6, 7). However, thus far this prediction has only been fulfilled in transfers among relatively closely related species within the order Diptera, as seen with the Drosophilaelements mariner, hermes, hobo, andminos (7-9).

We decided to probe the evolutionary limits ofmariner's ability to transpose in vivo by introducing it into Leishmania major, a human pathogen belonging to the flagellate order Kinetoplastida, one of the earliest branching eukaryotic lineages (10). Success here could also provide genetic methods to study processes of virulence and pathogenesis in leishmaniasis, a widespread tropical disease that can frequently be fatal and for which satisfactory vaccines or chemotherapy are lacking. Although methods for stable DNA transfection and expression of foreign genes are well established in Leishmania, nonhomologous insertion of DNA has not been observed in stable transfections of this diploid organism (11). Mobilization of marinerwould thus provide a powerful tool for insertional mutagenesis in this pathogen.

The 1.3-kb Mos1 mariner element fromDrosophila mauritiana contains a single open reading frame (ORF) encoding the transposase, flanked by 28–base pair (bp) inverted repeats (12). An intact mariner element was inserted in one Leishmania vector (pX63PAC-Mos1, Fig.1A) (13), and a helper plasmid was used to provide transposase. Leishmania and other trypanosomatid protozoa synthesize mRNAs by a trans-splicing mechanism, where a 39-nucleotide mini-exon is added to the 5′ end of every mRNA (14). Accordingly, the mariner transposase coding region was inserted in a Leishmania expression vector (pX63TKNEO-TPASE, Fig. 1A) (13) downstream of a trans-splice acceptor site because Drosophila genes lack these RNA signals. The two plasmids were then introduced into Leishmania major line +/Δ1 (15). Leishmania plasmids are maintained as stable episomes while under drug pressure (G418 and puromycin for the NEO and PAC markers, respectively) but are slowly lost during growth in the absence of selection (16).

Figure 1

Transposition of marinerin Leishmania. (A) Representation of the transposase expression plasmid (pX63TKNEO-TPASE), themariner donor plasmid (pX63PAC-Mos1), and the MosHYG donor plasmid (pX63PAC-MosHYG) (13). Coding regions (TKNEO, TPASE, PAC,HYG) are shown by black arrows, and the predicted mRNAs directed by the flanking elements present in the Leishmaniavectors are indicated by the light gray boxes. The plasmid ampicillin marker is shown by a narrow black box (Amp). The marinerelement is shown by a white box, flanked by larger black boxes marking the 5′ and 3′ inverted repeats (IR). X, Xho I; B, Bam HI, and H, Hind III. (B) Southern analysis. All lines derive from +/Δ1L. major (15), and all DNAs were digested with Hind III. Lanes: pTPASE, pX63TKNEO-TPASE DNA; pMos1, pX63PAC-Mos1 DNA; WT, +/Δ1 genomic DNA; TPASE1 and TPASE2, genomic DNA from independent pX63TKNEO-TPASE–transfected lines; TPASE1 + pX63PAC-Mos1, genomic DNA from 22 independent colonies containing both plasmids pX63TKNEO-TPASE and pX63PAC-Mos1. Arrows point to new mariner bands; differences in intensity may reflect transposition at various times during colony growth. The mariner hybridization probe was the 1.1-kb Cla I–Nhe I fragment from pBluescribe M13+/Mos1 (12). Size markers are in kilobases. (C) DNA sequence of mariner transpositions. Inverse PCR (19) was used to recover the new fragments evident in clones 7, 21, and 22 [(B), arrows], and their DNA sequence was determined. The Drosophila sequences present in theMos1 donor plasmid are in italics, the marinersequence is in the central gray box, and the TA dinucleotides marking the boundary between mariner and Leishmaniasequences are in bold type. Because 5′ and 3′ sequences were obtained independently, they possibly may not arise from the same insertion event.

Transfectant colonies were analyzed formariner transposition first by Southern (DNA) blot hybridization (17). Despite the lack of any selection for transposition, 5 of 22 colonies (23%) showed newmariner-hybridizing bands (Fig. 1B). No evidence of transposition was obtained in Southern blot analysis of 52 colonies containing only pX63PAC-Mos1 (18). The marinerinsertion site from several of the new fragments arising in the presence of transposase was obtained by inverse polymerase chain reaction (PCR) (19). Sequence analysis showed that they contained mariner, followed by a TA dinucleotide and sequences not present in the donor plasmid DNA (Fig. 1C) (19). Southern blot hybridizations with the newmariner-flanking sequences showed that they were ofLeishmania origin (20). Moreover, their fragment size had increased by 1.3 kb in the colony that gave rise to the PCR product (20), as expected for bona fide transposition.

The frequency of mariner insertion into a specific locus was measured for dihydrofolate reductase–thymidylate synthase (DHFR-TS). The parental +/Δ1 line used in the studies above is heterozygous, having one copy of DHFR-TSover a deletion allele, and we have shown previously (21) that one can select against DHFR-TS by plating these parasites in the presence of methotrexate (MTX) and thymidine (TdR). Surviving parasites undergo either loss-of-heterozygosity (LOH) or inactivating point mutations, and we anticipated thatmariner insertion into DHFR-TS would similarly inactivate it.

Plating of mariner-containing strains on MTX plus TdR yielded MTX-resistant (MTXr) colonies at a frequency of 1.2 × 10−4, as reported previously (21). Southern blot analysis showed that 39 colonies exhibited LOH, and 9 colonies retained a presumably alteredDHFR-TS allele [Fig. 2A or (20); line 7M9 shows LOH and the rest retain DHFR-TS sequences]. Of the non-LOH colonies, one (22M3) exhibited a DHFR-TSfragment of 7.1 kb, 1.3 kb larger than the wild-type 5.8-kb fragment (Fig. 2A), and rehybridization of the blot with a marinerprobe identified the same 7.1-kb fragment (Fig. 2B). Sequencing of the 22M3 mariner insertion confirmed that it had transposed intoDHFR-TS, into a TA dinucleotide located at position 532 within the DHFR coding region, and led to duplication of the target TA (Fig. 2C). Together with the results shown in Fig. 1C, it is evident that transposition in Leishmania proceeds by the characteristic mechanism of the mariner/Tc1 family, involving insertion into and duplication of an invariant target TA dinucleotide (22).

Figure 2

Insertional inactivation ofDHFR-TS by mariner. (A) Southern blot analysis of the DHFR-TS locus from colonies obtained after selection against DHFR-TS expression by plating of lines 7, 17, and 22 (Fig. 1B) on MTX plus TdR. WT, +/Δ1; the remaining lines are named after the original colony and the specific independent clone obtained after selection. Thus, line 22M3 is the third colony examined after plating of line 22. DNAs were digested with Eco RV and Hind III, and the probe was a Bgl II–Sty I 0.7-kb fragment of the coding region of the Leishmania donovani DHFR-TS gene (clone pDDHFR8; strain B2064). Size markers are in kilobases. (B) The probe from the blot in (A) was removed, and the blot was rehybridized with the mariner probe (see legend to Fig. 1B). In some colonies, the parental donor plasmid was retained. (C) Sequence of the mariner insertion in mutant 22M3 and comparison with the wild-type DHFR-TS sequence. The nucleotide coordinates for the DHFR-TS coding region are shown flanking the sequence, and the amino acid sequence is shown above. The TA insertion site and its transposition-generated duplication in line 22M3 are shown in bold. Single-letter abbreviations for the amino acid residues are as follows: A, Ala; E, Glu; I, Ile; K, Lys; L, Leu; Q, Gln; R, Arg; T, Thr; V, Val; and Y, Tyr.

Several of the random colonies lacking insertions inDHFR-TS showed new mariner-hybridizing fragments (7M8, 7M9, and 17M11; Fig. 2B), suggestive of transposition events involving other unselected loci within the Leishmaniagenome. Because of the discontinuation of G418 and puromycin selection, the parental plasmids were lost in some colonies (22M8 and 22M10; Fig.2B).

An estimate for the frequency of marinertransposition in Leishmania was calculated, (1/48) × (1.2 × 10-4), or about 2.5 × 10-6 for the insertional inactivation of a single allele. Because DHFR-TS spans 1.5 kb, and theLeishmania genome is about 50,000 kb and diploid, this leads to an estimated frequency for independent marinertranspositions of more than 10% per genome. This is in good agreement with the results obtained from Southern blots of random colonies (Figs. 1B and 2B). Future studies may permit more accurate measurement of the rate of transposition and the development of methods for increasing or regulating it.

We asked whether a modified mariner element carrying a drug resistance marker could be used to “trap” and thereby identify new genes within Leishmania. The MosHYG element contains a hygromycin resistance gene (HYG) inserted immediately adjacent to the mariner 28-bp inverted repeats, 5′ of the transposase (Fig. 1A). In this configuration, theHYG gene in MosHYG lacks Leishmania splice acceptor elements, contains several upstream in-frame stop codons, and is placed in an antisense orientation relative to mRNA processing signals present elsewhere in the pX63PAC vector (Fig. 1A). Thus, hygromycin resistance should only arise after transposition of MosHYG downstream of a Leishmania splice acceptor.

Transfections of 10 μg of pX63PAC-MosHYG DNA into cells containing pX63TKNEO-TPASE yielded 110 to 160 colonies when plated on puromycin media but no colonies when plated on hygromycin media, confirming that HYG is initially silent in the context of the donor plasmid. However, replating eight independent lines containing both plasmids on hygromycin media yielded colonies at frequencies ranging up to 10−3 (23). In contrast, selection of 21 cell lines containing only pX63PAC-MosHYG did not yield any hygromycin-resistant parasites, suggesting thatLeishmania lacks an endogenous transposase activity.

Transpositional activation of HYG should generate chimeric mRNAs containing the Leishmania mini-exon sequence, a variable amount of intervening 5′ untranslated sequence, and the MosHYG element (Fig. 3A). Reverse transcriptase–PCR (RT-PCR) with mini-exon– andHYG-specific primers was used to recover the predicted mRNA in several lines, and sequence analysis revealed the expected chimeric structure (24) (Fig. 3A). These findings were confirmed by Northern (RNA) blot analysis. When a probe derived from transposition event T3.6B was used, a 4.4-kb mRNA was observed in wild-type parasites, or in parasites bearing different transpositions. In contrast, a 6.9-kb mRNA was additionally found in the parental line T3.6 (Fig. 3, A and B), and, as expected, the increase in size (2.5 kb) was close to the size of the MosHYG element (2.4 kb). Similar results were obtained with the T3.5B probe (20) (Fig. 3A).

Figure 3

Use of a modified mariner element to trap expressed regions of the Leishmania genome. (A) Sequence of RT-PCR products corresponding to transposition events generating chimeric HYG-containing RNAs (24). The numbers indicate the specificLeishmania line (1.1, 1.2, 3.5, 3.6, 3.8, 6.2, and 8.1) obtained after plating colonies bearing pX63TKNEO-TPASE and pX63PAC-MosHYG on hygromycin plates; the letters A to D denote the specific PCR amplification product sequenced. The structure of the predicted chimeric transcripts is shown above the sequences and consists of the mini-exon (open box on left), Leishmaniasequences (dashed line), the 28-bp mariner inverted repeat (heavy black line), and the HYG resistance marker (gray box). For clarity, dashes were arbitrarily added to T6.2A, and the entire sequence of the products is not shown (nt, nucleotides). (B) Northern blot analysis of total RNAs from the +/Δ1 (WT) and several hygromycinrlines described in (A). The hybridization probe was prepared by inverse PCR and corresponded to the T3.6B product (A), which arose from line 3.6 (40). Similar results were obtained with the T3.5B product, which arose from line 3.5 (20). Size markers are in kilobases.

Most of the trapped Leishmania sequences did not show matches in database searches, although one (T1.2D) showed insertion into the 5′ untranslated region of the DHFR-TSlocus (Fig. 3A). Further analysis showed that in this colony,mariner had transposed into the donor plasmid pX63PAC-MosHYG, which contains DHFR-TS flanking sequences to drive expression of the PAC marker (25). Because the DHFR-TS splice acceptor site was used normally (26), and Northern blot analyses revealed the expected increases in mRNA size from insertion of the MosHYG elements, we conclude that mariner insertion does not interfere with processing of endogenous mRNAs.

In summary, we have shown that mariner can transpose efficiently in Leishmania and have used it as an insertional mutagen and to trap new Leishmania genes. Classical genetic studies of Leishmania as well as of other medically important organisms such as trypanosomes and several pathogenic fungi are hampered because these parasites are diploid and often lack an experimentally manipulable sexual cycle. Thus, reverse genetic approaches are particularly important. The development of amariner-based heterologous transposon system should prove a significant addition to the array of tools available for dissecting important aspects of Leishmania biology, such as virulence and pathogenesis.

Our data reinforce the impression thatmariner and related elements are autonomous and able to cross distant species boundaries in vivo. Many workers have suggested that mariner could be adapted for use as a wide-range transformation vector (2, 3, 6, 7). Previously, heterologousmariner/Tc1 transposition had been observed, but only within members of the same taxonomic order (7, 8). Our findings now extend the range to different kingdoms, separated by an evolutionary distance of probably more than 1 billion years. To our knowledge, this is the widest evolutionary distance yet shown to be traversed by transposable elements in vivo and suggests that mariner's potential utility may be comparably broad. Thus, in addition to its role in shaping eukaryotic genomes (3, 27), this parasitic DNA can now be applied toward probing the genomes of human parasites and, conceivably, many other eukaryotes.

  • * To whom correspondence should be addressed. E-mail: beverley{at}borcim.wustl.edu

  • Address after 1 July 1997: Department of Molecular Microbiology, Washington University School of Medicine, St. Louis, MO 63110, USA.

REFERENCES AND NOTES

View Abstract

Navigate This Article