Unexpected Stability of mariner Transgenes in Drosophila

Corresponding author: Daniel L. Hartl, Harvard University, 16 Divinity Ave., Cambridge, MA 02138., dhartl{at}oeb.harvard.edu (E-mail)

Communicating editor: J. A. BIRCHLER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

A number of mariner transformation vectors based on the mauritiana subfamily of transposable elements were introduced into the genome of Drosophila melanogaster and examined for their ability to be mobilized by the mariner transposase. Simple insertion vectors were constructed from single mariner elements into which exogenous DNA ranging in size from 1.3 to 4.5 kb had been inserted; composite vectors were constructed with partial or complete duplications of mariner flanking the exogenous DNA. All of the simple insertion vectors showed levels of somatic and germline excision that were at least 100-fold lower than the baseline level of uninterrupted mariner elements. Although composite vectors with inverted duplications were unable to be mobilized at detectable frequencies, vectors with large direct duplications of mariner could be mobilized. A vector consisting of two virtually complete elements flanking exogenous DNA yielded a frequency of somatic eye-color mosaicism of 10% and a frequency of germline excision of 0.04%. These values are far smaller than those observed for uninterrupted elements. The results imply that efficient mobilization of mariner in vivo requires the presence and proper spacing of sequences internal to the element as well as the inverted repeats.

THE transposable element mariner is enumerated among a handful of sequences that have practical utility in serving as vectors for transformation in a variety of organisms (HARTL et al. 1997 ; PLASTERK et al. 1999 ; ATKINSON et al. 2001 ; HANDLER 2001 ; HARTL 2001 ). For mariner, the organisms successfully transformed so far include various bacteria, such as Escherichia and Mycobacterium (RUBIN et al. 1999 ), Helicobacter (GUO and MEKALANOS 2001 ), Vibrio (CHIANG and MEKALANOS 2000 ), and Neisseria (PELICIC et al. 2000 ). They also include various species of drosophilid insects (LIDHOLM et al. 1993 ; LOHE and HARTL 1996B ; HORN et al. 2000 ), as well as an impressive list of nondrosophilids including mosquitoes (COATES et al. 1998 ; MOREIRA et al. 2000 ), silkworm (WANG et al. 2000 ), and housefly (YOSHIYAMA et al. 2000 ). The mariner element has proven successful in transforming eukaryotes as different as the protozoan Leishmania (GUEIROS-FILHO and BEVERLEY 1997 ) and the vertebrates chicken (SHERMAN et al. 1998 ) and zebrafish (FADOOL et al. 1998 ).

The wide variety of organisms that can be transformed by mariner vectors is not surprising in view of its extraordinary host range and ability to undergo horizontal transmission. Although the element was initially discovered in Drosophila mauritiana (JACOBSON et al. 1986 ) and shortly thereafter in Zaprionus tuberculatis (MARUYAMA and HARTL 1991A ) as well as Hyalophora cecropia (LIDHOLM et al. 1991 ), its broad taxonomic distribution began to become clear only with the work of ROBERTSON 1993 , ROBERTSON 1995 and ROBERTSON and MACLEOD 1993 who found that >15% of 400 insect species examined had one or more diverse subfamilies of mariner elements present in their genome. Soon thereafter mariner elements were found in centipedes and mites (ROBERTSON and MACLEOD 1993 ), nematodes (SEDENSKY et al. 1994 ), and planaria (GARCIA-FERNANDEZ et al. 1993 , GARCIA-FERNANDEZ et al. 1995 ). Mariner elements have also been found in humans and other vertebrates (AUGE-GOUILLOU et al. 1995 ; MORGAN 1995 ; OOSUMI et al. 1995 ; ROBERTSON et al. 1996 ; SMIT and RIGGS 1996 ), as well as in plants (JARVIK and LARK 1998 ).

The canonical mariner element denoted Mos1 is 1286 bp in length and includes 28-bp inverted repeats. The element contains an uninterrupted open reading frame encoding an active transposase of 345 amino acids (MEDHORA et al. 1988 ; MARUYAMA et al. 1991 ). The mariner element originally discovered, called peach, is also a full-length element. It differs from Mos1 at 11 nucleotide sites. The peach element retains an open reading frame and is able to transpose, but the peach transposase differs from that of Mos1 at 4 amino acid sites and is inactive (MARUYAMA et al. 1991 ; MEDHORA et al. 1991 ).

Both peach and Mos1 have been used as transformation vectors in eukaryotes, with the active Mos1 transposase used for mobilization into the target genome. The elements feature a number of convenient unique restriction sites, including SacI at nucleotide position 787–792 (numbering as in JACOBSON et al. 1986 ), SalI at position 349–354, SphI at position 987–992, and ClaI at position 63–68 (this site is present in Mos1 but not in peach).

In our original transformation experiments utilizing a peach vector with 11.9 kb of exogenous DNA inserted at the SacI site (LIDHOLM et al. 1993 ), we observed that genomic insertions were remarkably stable in the presence of active transposase (LOHE et al. 1995 ). Genomic insertions of another vector with only 4.5 kb of exogenous DNA inserted at the SacI site were also unexpectedly stable (LOHE et al. 1995 ). Both kinds of genomic insertions can be mobilized, but at low frequency (LOHE and HARTL 1996C ).

To investigate whether mariner vectors with insertions of exogenous DNA at other unique cloning sites show levels of genetic stability comparable to insertions at the SacI site, we carried out the experiments reported in this article. We studied vectors with insertions at the SalI, SphI, and ClaI sites. The inserted DNA was variously a mini-white marker (THUMMEL and PIRROTTA 1992 ), a yellow marker (PATTON et al. 1992 ), or an eye-specific enhanced green fluorescent marker responsive to the transcription factor Pax-6 (BERGHAMMER et al. 1999 ; HORN et al. 2000 ). We also studied novel types of composite vectors in which the exogenous DNA was flanked by two essentially complete mariner elements in any of three possible orientations. In all cases with exogenous DNA inserted into the SalI, SphI, or ClaI site, we find that genomic insertions are largely refractory to mobilization by the active Mos1 transposase. Somatic eye-color mosaicism is not observed, nor are progeny resulting from germline excision detected. Low levels of excision can, however, be detected by means of PCR amplification. With regard to the composite vectors, only genomic insertions with the flanking mariner elements in the same orientation (the "tail-to-head" configuration) were susceptible to mobilization by the Mos1 transposase. We observed a rate of somatic mosaicism of 1 in 10 and a rate of germline excision of 1 in 2000. Although still substantially smaller than that rate of somatic mosaicism and germline excision observed with the peach element alone (100 and 14%, respectively), nevertheless the composite tail-to-head vector has a level of genomic instability far greater than any of the simple insertion vectors. These results are discussed in light of their implications for the use of mariner as a tool in genetic analysis and in the creation of transgenic organisms.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Two types of transformation vectors were investigated. Simple insertion vectors consist of a single copy of mariner into which a marker gene was inserted at any one of a number of unique restriction sites. Composite vectors consist of a duplication of mariner sequence with a marker gene inserted at the junction of the duplicated regions.

Simple insertion vectors:
Vectors were created with insertions in the mariner SalI site (begins at nucleotide 349), the SphI site (begins at nucleotide 987), or the ClaI site (begins at nucleotide 63). To prepare these vectors, the mariner element denoted peach was subcloned from the plasmid pJJ1 (JACOBSON et al. 1986 ) as an EcoRV-SpeI fragment and subcloned into pBlueScriptII KS, yielding the plasmid pPch. Polylinker sequence flanking the EcoRV-SpeI peach fragment was modified to remove the ApaI and SalI sites as well as to introduce flanking EcoRI sites. Two adaptors were prepared, one by annealing oligonucleotides Kpn/Hind-S (5'-CGCGGCCGCGAATTCA-3') with Kpn/Hind-L (5'-AGCTTGAATTCGCGGCCGCGGTAC-3'), and the other by annealing Spe/Not-1 (5'-CTAGAATTCAGATCT-3') with Spe/Not-2 (5'-GGCCAGATCTGAATT-3'). These adaptors were ligated with the HindIII-SpeI peach-containing fragment of pPch and inserted into pBlueScriptII KS cleaved with KpnI and NotI. The resulting derivative of pPch was designated pPch-RR.

The hsp70:mini-white cassette was subcloned as an XbaI-PstI fragment from PlwB (LIDHOLM et al. 1993 ) into pGem-3Zf(+). A trimolecular ligation was then used to create a duplication of polylinker at the ends of the hsp70:mini-white fragment by excising the hsp70:mini-white by digestion with BamHI and SphI and ligating with both an AatII-SphI fragment and a BamHI-AatII fragment from pGem-3Zf(+). In the resulting plasmid, called hsp/w-BSBS, the hsp70:mini-white cassette is flanked by both BamHI and SphI sites. This hsp70:mini-white cassette was excised from hsp/w-BSBS with BamHI and inserted into the SalI site of the peach element in pPch-RR after half filling the cohesive ends with Klenow polymerase. This yielded the insertion vector designated SalW. (Both orientations of the hsp70:mini-white were recovered, but only one was studied further.) Similarly, the hsp70:mini-white cassette was excised from hsp/w-BSBS with SphI and inserted into the SphI site of the peach element in pPch-RR. This yielded the insertion vector designated SphW. (Again both orientations of the hsp70:mini-white were recovered, but only one was studied further.)

To obtain the insertion of hsp70:mini-white into the ClaI site of mariner, we had to modify the sequence of the peach element because, unlike the Mos1 element, the peach element lacks a ClaI site at position 63. Accordingly, nucleotides 7–1280 of the peach element were excised from pPch-RR using the restriction enzyme BsrGI, and this fragment was replaced with the corresponding fragment from the plasmid pMos1 (MEDHORA et al. 1988 ). This operation completely converted the sequence of the peach element in pPch-RR into that of Mos1, yielding the plasmid pMos-RR. In the next step, the kanamycin-resistant element in the plasmid pUC4K was excised with BamHI and inserted in the ClaI site of pMos-RR after rendering the cohesive ends blunt with Klenow fragment, which restores the BamHI sites flanking the kanamycin-resistant insert. Replacing the kanamycin-resistance cartridge with the hsp70:mini-white using BamHI restriction sites resulted in the vector designated ClaW. The vector ClaY, which carries a yellow marker gene in the ClaI site, was created from pMos-RR by excising a 5.2-kb yellow-containing SalI fragment from pCar-y (PATTON et al. 1992 ) and ligating it into the ClaI site of pMos-RR after filling the cohesive ends with Klenow fragment to produce blunt ends.

The vector designated SalG contains a marker gene inserted in the SalI site of the Mos1 mariner element. The marker gene, 1.3 kb in length, consists of a coding sequence for an enhanced green fluorescent protein driven by an artificial promoter that is eye specific and responsive to the evolutionary conserved transcription factor Pax-6. The SalG vector is described in detail as pMos{3xP3-EGFP} in HORN et al. 2000 .

Composite vectors:
The vector designated SemiComp contains a direct duplication of the SalI-SphI fragment of peach with an hsp:mini-white inserted at the junction of the duplicated fragments. This vector was created from the simple insertion vectors SalW and SphW, using the ApaI restriction site in the hsp:mini-white. The 5' part of SphW was isolated as a KpnI-ApaI fragment and used to replace the 5' part of SalW. In the resulting vector SemiComp the hsp:mini-white sequence is flanked by a direct duplication of the SalI-SphI segment of the peach element.

Three types of composite vectors containing almost complete duplications of peach were also created. The first step in generating these vectors was the introduction of mutations in the inverted repeats destined to flank the marker gene in the composite vector. For this purpose a peach element was created in which a PstI restriction site was introduced to replace the first nucleotide of the 5' inverted repeat and the flanking TA duplication. This 5'-mutated element was amplified from the plasmid pJJ1 (JACOBSON et al. 1986 ), using the primers 5'-mut (5'-CTGCAGGTGTACAAGTAGGGAATGTCGG-3'), which is specific for the 5'-terminal inverted repeat, and 3'-pch (5'-TTGTAGCGTTACCTAGCGTC-3'), which is homologous to the sequence of the intron of the white gene adjacent to the 3' end of the peach element in pJJ1. Similarly, the 3'-mutated element was amplified with the primers 3'-mut (5'-CTGCAGGTGTACAAGTATGAAATGTCGT-3') and 5'-pch (5'-GTTTTGGCACAGCACTTTGTG-3'). Both mutated elements were cloned in the vector pCRII (Invitrogen, Carlsbad, CA). The resulting clones were stable only in the SURE strain of Escherichia coli (Stratagene, La Jolla, CA), which was used in all subsequent experiments. Before proceeding further, the integrity of the PCR products was verified by DNA sequencing.

The 5'-mutated and 3'-mutated peach elements were used to create three kinds of composite vectors, designated TT (tail-to-tail), HH (head-to-head), and TH (tail-to-head), in which the duplicated peach elements are present in the orientations 5'-3'/3'-5' (TT), 3'-5'/5'-3' (HH), and 5'-3'/5'-3' (TH). In all cases the two innermost inverted repeats are the mutated ends and the outer inverted repeats are those of the unmutated peach element. In each type of vector an hsp70:mini-white gene was inserted between the innermost pair of inverted repeats. The peach inverted repeats directly adjacent to the mini-white gene are modified and not susceptible to excision by the mariner transposase (LOHE et al. 2000 ). The outermost inverted repeats are intact and could, in principle, be used by transposase to mobilize the entire vector.

To create the TT vector, the 3'-end mutated peach element was excised from pCRII with EcoRI and PstI and the fragments were ligated in a reaction including a PstI fragment containing hsp70:mini-white derived from PlwB (LIDHOLM et al. 1993 ) and plasmid pSP72 (Promega, Madison, WI) that had been linearized with EcoRI. The HH configuration was created similarly, using the 5'-end mutated peach element from pCRII.

Construction of the TH vector made use of both the EcoRI-PstI fragment containing the 5' mutation and the EcoRI-PstI fragment containing the 3' mutation. These were ligated together with plasmid pSP72 that had been linearized with EcoRI. The only peach-containing clones recovered were those containing the tandemly repeated elements in the tail-to-head configuration, yielding the plasmid pSP-TH. Then the PstI fragment containing the hsp70:mini-white cassette from the plasmid PlwB (LIDHOLM et al. 1993 ) was inserted into the PstI site of pSP-HT, creating the TH vector.

Insertion of the vectors into the genome:
The original SacW vectors were introduced into the germline by mariner-mediated transposition using Mos1 as a helper plasmid (LIDHOLM et al. 1993 ; LOHE et al. 1995 ). In the present analysis, the SalG vectors were also introduced into the germline via mariner-mediated transformation as described in HORN et al. 2000 . The other vectors were introduced using the Hermes transformation system (O'BROCHTA et al. 1996 ) by injection into embryos of genotype w¹¹¹⁸ or y w. Each of the vectors of interest is flanked by unique EcoRI sites in the plasmid in which they reside. The Hermes vectors were constructed from the plasmid pBSHermes w⁺ (O'BROCHTA et al. 1996 ) by removing the white-bearing EcoRI fragment in this plasmid and replacing it with an EcoRI fragment bearing the mariner vector and marker gene of interest. The resulting plasmid was employed in germline transformation using a helper plasmid pBCHSHH1.9 (O'BROCHTA et al. 1996 ), which bears a Hermes transposase open reading frame under the control of a heat-shock promoter. Concentrations of plasmid and helper were both 0.5 µg/µl, and the methods were those described in LOZOVSKAYA et al. 1996 .

Excision assays:
Genetic experiments to estimate the in vivo rates of somatic and germline excision of the inserted vectors were carried out as described in LOHE et al. 1995 , LOHE et al. 1997 .

Molecular methods:
The polymerase chain reaction (PCR) was used to detect excision products at the molecular level, using primers flanking the EcoRI insertion site in the Hermes vector, namely 5'-ATAAATGCTGTGCCTCTCTA-3' and 5'-ATTGTTTGTAGTATTGC-3'. These primers amplify excision products to yield a PCR product of 475 bp (51 bp of Hermes plus 42 bp of Drosophila sequence on the 5' side of the 35-bp polylinker and 264 bp of Hermes plus 83 bp of Drosophila sequence on the 3' side). The PCR conditions were 95° for 1 min, then 25 cycles of 95° for 20 sec, 50° for 90 sec plus 1 additional second for each cycle, and 72° for 90 sec plus 4 additional seconds for each cycle. The resulting PCR products were stored at 4°. The PCR was carried out in a PTC100 machine from MJ Research (Waltham, MA). Southern blotting and DNA sequencing were carried out as previously described (NURMINSKY et al. 1998 ).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We examined a number of possible transformation vectors on the basis of the transposable element mariner. Some of these were simple insertion vectors having exogenous DNA inserted into unique restriction sites in the element, and others were composite vectors in which a large region (in most cases virtually all) of the mariner element was duplicated on either side of the exogenous DNA. The vectors differed in the nature of the exogenous DNA, which included a 4.5-kb mini-white gene, a 5.2-kb yellow gene, or a 1.3-kb green fluorescent protein (GFP) cassette. In some vectors the direction of transcription of the marker gene was in the same orientation as that of the mariner open reading frame and in other cases in the reverse orientation. Most of the vectors were introduced into the genome of D. melanogaster using the Hermes transformation system. At first we tried the P-element system, finding that many of the simple insertion vectors were genetically unstable when subcloned into standard P-element vectors, yielding deletions and rearrangements (data not shown). We also tried the hobo transformation system (BLACKMAN et al. 1989 ) but found, as others have, that many of the standard laboratory strains produce low levels of functional hobo transposase, which compromised the assays for the excision of mariner vectors (data not shown). Finally we turned to the Hermes system (O'BROCHTA et al. 1996 ), which yielded genetically stable subclones, a high efficiency of transformation, and genetic stability in the germline of the laboratory strains we used for the analysis. In most cases we recovered more germline transformants than were actually analyzed for vector excision. About one-third of the Hermes transformants had multiple insertions as evidenced by Southern blots and/or in situ hybridization (data not shown). All of the results described below were obtained from transformants that carried only one Hermes insertion.

The vectors were tested for their ability to be excised by the functional mariner transposase produced by the immobile element Mr182 as described in LOHE et al. 1995 , LOHE et al. 1997 . Mr182 is a P[hsp70::Mos1, ry⁺]-182 construct inserted in chromosome 2; the hsp70::Mos1 sequence has a dual promoter in which the heat shock 70 promoter (hsp70) is fused to the Mos1 promoter at nucleotide position 58–59 (LOHE et al. 1995 ). The dual promoter has high activity even in the absence of heat shock (LOHE et al. 1995 ). Rates of somatic excision and germline excision of mariner elements are highly correlated with each other and with the rate of transposition (GARZA et al. 1991 ; LOHE et al. 1995 , LOHE et al. 1997 ; LOHE and HARTL 1996A , LOHE and HARTL 1996C ). The baseline rate of somatic excision of the peach element in the presence of Mr182 is 100 percent—which is to say that every fly shows somatic mosaicism of its eye-color phenotype (GARZA et al. 1991 ; LOHE et al. 1997 , LOHE et al. 2000 ). The baseline rate of germline excision of peach is 14.7% (LOHE et al. 2000 ); this value reflects the proportion of revertant chromosomes in the mosaic germline.

Simple insertion vectors:
The structure of each of the insertion vectors is diagrammed in Fig 1, which shows the insertion site and orientation of the exogenous DNA. In all vectors except for ClaW, ClaY, and SalG, the mariner element is the peach element (JACOBSON et al. 1986 ); the mariner element used for the vectors ClaW, ClaY, and SalG is Mos1 (GARZA et al. 1991 ). The peach and Mos1 elements differ at 11 nucleotide sites, including one that creates the ClaI site in Mos1. The SacI vector exists in two forms with either an 11.9-kb white insertion or a 4.5-kb mini-white insertion. Results with these vectors have been reported previously (LIDHOLM et al. 1993 ; LOHE and HARTL 1996C ). Among the transformants, the level of somatic mosaicism is 1%, which is lower than the baseline value by two orders of magnitude.

View larger version (23K): In this window In a new window Download PPT slide   Figure 1. Molecular structures of simple insertion vectors. Inserted exogenous DNA is not drawn to scale.

The SalG transformants could not be scored for somatic mosaicism because small nonfluorescent patches, if there were any, in an otherwise highly fluorescent eye would probably not be detectable. The rates of somatic excision for all of the simple insertion vectors in Table 1 are at least 400 times smaller than the baseline values for peach. Experiments with comparable numbers of progeny were carried out to estimate the rate of germline excision, but no germline excision was detected for any of the vectors. For seven independent SalG insertions, we examined progeny for nonfluorescent eyes resulting from germline excision and observed 0/831. Based on the binomial distribution, this value defines an upper 95% confidence bound for the frequency of germline excision of the SalG vector of 0.36%, which is far smaller than the baseline value of 14.7% observed for the peach element alone (LOHE et al. 2000 ).

As Table 1 indicates, for most of the simple insertion vectors the rate of somatic excision is too low to be detected phenotypically as eye-color mosaicism among a few thousand progeny. Nevertheless, for some of the vectors, somatic excision can readily be detected at the molecular level in each individual fly. To detect the excision footprints we used PCR with primers flanking the EcoRI insertion site of the mariner vector in the Hermes transposon. Excision of a simple insertion vector would result in a predicted PCR product of 475 bp. A fragment of this size was prominent in the PCR product of each of 5–10 nonmosaic flies carrying either SalW or SphW, but not ClaW and ClaY. In the latter cases the excision product was virtually undetectable, either with amplification of DNA from single flies or from DNA isolated from homogenates of 20 flies. A sample of 10–12 PCR products from apparent excisions of the other vectors was isolated and the relevant region sequenced to ascertain whether they exhibited the characteristic footprint of mariner excision. In each case the former location of the peach vector was now occupied by a sequence consisting of the TA duplication flanking peach plus one to three nucleotides from either the 5' inverted repeat (5'-CCA-3') or the 3' inverted repeat (5'-TGA-3'). These sequences are typical of excision mediated by the mariner transposase (BRYAN et al. 1990 ). (It should be noted that the PCR tests were not performed with the SalG transformants because this vector was introduced into the genome via mariner-mediated transformation rather than via Hermes transformation, and so the genomic sequences flanking the SalG transformants are unknown.)

Composite vectors:
The low rate of excision of the simple insertion vectors might result from a tight size constraint whereby the Mos1 transposase cannot efficiently mobilize elements substantially longer than the 1286-bp length of peach and Mos1. Alternatively (or perhaps in addition) mariner may contain two or more internal transposase-binding sites whose proper spacing is necessary for efficient mobilization. To examine this possibility, various types of composite transposons that contain duplications of some or all internal sequences were constructed from the peach element.

The structure of the composite vectors is shown in Fig 2. The SemiComp vector contains a direct duplication of the SalI-SphI region, between which the exogenous mini-white DNA is inserted. The fully composite vectors (TT, HH, and TH) have two almost complete peach elements flanking the exogenous mini-white DNA. These vectors differ in the orientation of the peach elements. In the TT vector, the 3' ends of peach abut the exogenous DNA, and in the HH vector the 5' ends of peach abut the exogenous DNA. In the TH configuration, both peach elements are in the same orientation, with the 3' end of the upstream element and the 5' end of the downstream element abutting the exogenous DNA. In each of the fully composite vectors, the inverted repeat adjoined with the mini-white sequence has been mutated by deletion of the terminal nucleotide and the flanking TA, which have been shown to be critical for efficient recognition and cleavage by the mariner transposase (LOHE et al. 2000 ).

View larger version (24K): In this window In a new window Download PPT slide   Figure 2. Molecular structures of composite vectors. Inserted exogenous DNA is not drawn to scale.

For transformants carrying each of the composite vectors, the frequency of somatic mosaicism is shown in Table 2. Although the two vectors with inverted duplications (TT and HH) show no evidence of somatic excision, the levels of somatic excision observed for the vectors with direct duplications (SemiComp and TH) are at least an order of magnitude greater than that of the simple insertion vectors (Table 1). Furthermore, the TH vector, which has all of the internal peach sequence duplicated, exhibits a significantly greater level of somatic excision than that of the SemiComp vector, which has only the SalI-SphI fragment duplicated (P = 0.03, Fisher's exact test). Although most of the eye-color mosaics carrying the TH construct have white (vector excision) patches on an otherwise red background, many of them also have one or more patches of eye facets that are exceptionally dark red. We attribute these to dosage effects of the mini-white marker resulting from somatic transposition of the TH vector, yielding cell lineages with two copies of the vector. In some crosses as many as 12.5% of the somatic mosaics showed such dark patches.

PCR analysis was carried out on flies carrying the TT and HH constructs, as well as on nonmosaic flies carrying the TH vector. For the TT and HH constructs, no PCR product of the size predicted from vector excision was detected. This result implies that excision of the TT and HH constructs occurs at an extremely low rate, if it occurs at all, which is consistent with previous results that the 5' inverted repeat and the 3' inverted repeat are not equivalent or interchangeable with respect to transposase recognition and cleavage (LOHE et al. 2000 ). PCR products arising from apparent excision were readily detected in the TH-bearing flies, however, and as expected their sequences contained the characteristic footprints of excision mediated by the Mos1 transposase. A sample of these footprints, each derived from a different animal, is shown in Table 3.

Although the frequency of somatic mosaics for the TH vector is 10-fold less than the baseline value for peach, it nevertheless suggested that the frequency of germline excision might be high enough to be detected in relatively small experiments. This proved to be the case. In crosses of males carrying an X-linked TH insertion with attached-X females, among 4515 male progeny we observed two revertants resulting from germline excisions. The estimated frequency of germline excision of the TH vector is thus 0.04%.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In this study we examined vectors with insertions of exogenous DNA into any one of several unique restriction sites (SalI, SphI, and ClaI) within the mariner, peach, or Mos1 elements, motivated by previous findings that vectors with insertions into the SacI site had a drastically impaired ability to be mobilized by the Mos1 transposase (LIDHOLM et al. 1993 ; LOHE et al. 1995 ). We also examined composite vectors with direct or inverted duplications of part or virtually all of the peach element. The exogenous DNA consisted of a 4.5-kb mini-white sequence (SalW, SphW, ClaW, and all of the composite vectors), a 5.2-kb yellow sequence (ClaY), or a 1.3-kb sequence encoding GFP (SalG). The vectors were introduced into the genome of D. melanogaster either by Mos1-mediated transformation (SalG) or by Hermes transformation (all other vectors). Transformed lines containing single insertions were then tested for somatic and germline stability in the presence of the Mos1 transposase source Mr182.

One important finding is that the simple insertion vectors have levels of somatic and germline excision that are at least 100-fold lower than the baseline level of uninterrupted peach elements. Low-level excision does take place, as shown by the analysis of PCR products from DNA of individual flies, which exhibited the characteristic sequence footprints of excision mediated by the mariner transposase. An exception should be noted for the ClaW and ClaY vectors, in which the expected PCR product was undetectable in standard gel assays.

Mobilization of some of the composite vectors was more efficient. Vectors containing a direct duplication of either the SalI-SphI fragment or almost the entirety of the element (except the 3' nucleotide and the flanking TA duplication) were the most efficiently mobilized. The composite vector TH was mobilized with an efficiency significantly greater than that of the partial duplication. The frequency of excision of the TH vector, in which the flanking peach elements are almost completely intact, was 10-fold greater than that of any of the insertion vectors, although still 10-fold smaller than that of the peach element itself. The TH vector also yielded a frequency of germline excision of 0.04%, sufficiently high to be detected phenotypically in experiments of typical size. Mobilization of the fully composite vectors required that the two peach elements be in the same orientation. Those with the flanking peach elements in inverted orientation gave no evidence for excision that could be detected either phenotypically or by PCR amplification.

Size of vector vs. insertion site:
The experiments suggest that overall size alone is not the primary cause of the impaired ability of the vectors to be mobilized. A simple insertion vector carrying 4.5 kb of exogenous DNA in the SacI site is not detectably more prone to mobilization than is one carrying 11.9 kb, and a vector with 1.3 kb inserted into the SalI site is not markedly more mobile than one carrying 4.5 kb in this site. The most efficiently mobilized composite vector we studied was the TH composite transposon, which has an overall size of 7.1 kb. This is 4.5 kb larger than the 2.66-kb SalG simple insertion vector. On the other hand, in the size range of 1.3- to 5.2-kb inserts that we have studied, while overall size is not the primary determinant of mobility, it may very well play some role in affecting the efficiency of mobilization. The TH composite exhibits a frequency of eye-color mosaicism of 10% and a germline excision rate of 0.04%. These values are still at least an order of magnitude smaller than the baseline values for peach and may to some extent reflect the 7.1-kb size of the vector in comparison with the 1.3-kb size of peach alone. In this context it should be noted that insect mariner elements found in natural populations that are closely related to Mos1 are all very close to 1.3 kb in size (CAPY et al. 1991 ; MARUYAMA and HARTL 1991B ; BRUNET et al. 2001 ).

Our results are most easily explained by supposing that efficient mariner mobilization requires the presence and correct spacing of certain internal sequences of the element relative to each other and to the inverted repeats. Insertion of exogenous DNA at almost any internal site would therefore impair the ability to be mobilized, although our data derive only from insertions at the ClaI, SalI, SacI, and SphI sites. The finding that composite vectors can be mobilized 10-fold more efficiently than simple insertion vectors is consistent with this interpretation. Some of the key internal sequences may be in the SalI-SphI fragment duplicated in the semicomposite vector, although the fully composite TH vector shows significantly greater mobilization, which may indicate that important internal sequences are also present outside the SalI-SphI region. The orientation of the internal sequences is evidently as important as their spacing, since we cannot detect mobilization of composite vectors that have their flanking peach elements in inverted orientation.

The hypothesis that essential internal sequences with the proper spacing and orientation are necessary for efficient mobilization is also consistent with recent results on the mobility of peach elements containing internal deletions. Analysis of 20 deletions revealed at least three internal regions whose integrity and spacing was necessary for efficient mobilization. These sequences are present within the regions bounded by nucleotides 229–586, 735–765, and 939–1066 (LOHE and HARTL 2001). The first of these regions contains the SalI site and the third the SphI site. We also found that mobilization was significantly impaired by a single nucleotide mutation at position 993 and by a double mutation at positions 161 and 179 (LOHE and HARTL 2001). How these internal sequences contribute to mariner mobilization is not known. Two obvious possibilities are that they are internal transposase-binding sites or that they are needed to form a secondary structure that is necessary for efficient recognition and cleavage. These models are of course not mutually exclusive.

In vivo results vs. in vitro results:
Our results on the impaired mobility of simple insertion vectors in vivo contrast with in vitro studies reporting that a Mos1 vector consisting of nucleotides 1–64 and nucleotides 1253–1286 flanking a 1.1-kb kanamycin-resistance gene can transpose as efficiently as a complete Mos1 element with the 1.1-kb kanamycin-resistance gene inserted into the SacI site (TOSI and BEVERLEY 2000 ). Two issues make the in vitro results difficult to compare with the in vivo results. The first is that the in vitro results are based on comparative data. Because the mini-Mos1 vector was compared only with a Mos1 vector having a SacI insertion, it is unclear how the levels of transposition compare with those that would be found in vitro for Mos1 alone. The second complication is that the in vitro experiments exploit the kanamycin-selection system, so levels of transposition could be detected that are far smaller than can be observed phenotypically. Most of our simple insertion vectors do show levels of somatic excision that can readily be detected by PCR amplification, though not phenotypically, so perhaps the in vitro and the in vivo results are not as discordant as they first appear. On the other hand, there may be real differences in transposition efficiency for any of several reasons. For example, in typical in vitro systems, the concentration of transposase protein relative to vector DNA may be quite different from what it is in vivo. Furthermore, an in vitro system is an assay for transposition from purified DNA, whereas an in vivo system is an assay for transposition from chromatin.

Comparison with other transformation systems:
Other systems for germline transformation based on insect transposable elements include the P element (RUBIN and SPRADLING 1982 ), hobo (BLACKMAN et al. 1989 ; LOZOVSKAYA et al. 1996 ), Hermes (O'BROCHTA et al. 1996 ), Tc1 (IVICS et al. 1997 ), Minos (LOUKERIS et al. 1995 ), and piggyBac (HANDLER et al. 1998 ; LOBO et al. 1999 ). In these systems the transposase appears to be efficient at mobilizing vectors that contain exogenous DNA of 1–5 kb at any of a number of internal sites. In terms of delivery of exogenous DNA into the germline, the Mos1 transformation system seems to be comparable to the others. The Mos1 system has been used to transform protozoans (GUEIROS-FILHO and BEVERLEY 1997 ), various species of insects (LIDHOLM et al. 1993 ; LOHE and HARTL 1996B ; COATES et al. 1998 ; HORN et al. 2000 ; MOREIRA et al. 2000 ; WANG et al. 2000 ; YOSHIYAMA et al. 2000 ), and vertebrates (FADOOL et al. 1998 ; SHERMAN et al. 1998 ). The efficiency of germline transformation is of the same order as that of other systems, even when the vectors are simple insertion vectors with exogenous DNA inserted at the ClaI site (GUEIROS-FILHO and BEVERLEY 1997 ), the SalI site (HORN et al. 2000 ), or the SacI site (LIDHOLM et al. 1993 ; LOHE and HARTL 1996B ). Indeed, HORN and WIMMER 2000 found that, when the SalI-NruI region of Mos1 (nucleotides 350–950) was deleted and replaced with the 1.3-kb GFP marker, the efficiency of Mos1-mediated germline transformation was increased relative to that observed with the SalG vector.

On the other hand, the Mos1 system manifests relative genetic stability of the vectors once they have been integrated into the germline. We are unaware of comparable findings published for other transformation systems. For mariner, the efficiency of mobilization of simple insertion vectors integrated into the genome of D. melanogaster is at least two orders of magnitude smaller than that of elements with no insert. The apparent discrepancy between relative efficiency of germline transformation compared with relative inefficiency of mobilization of the integrated vectors may relate to a difference between mobilization of elements that are present in molecules of plasmid DNA and those that are present in chromatin.

We do not know whether the relative stability of integrated simple insertion vectors is a characteristic of organisms other than D. melanogaster. It is possible that something peculiar about the chromatin of D. melanogaster does not occur in other organisms, which dramatically reduces the ability of simple insertion vectors to be mobilized once they have become integrated.

	FOOTNOTES

¹ These authors contributed equally to this work.

	ACKNOWLEDGMENTS

We thank Daniel De Aguiar and Javare Gowda Nagaraju for help in carrying out some of the experiments reported here. This work was supported by grants from the National Institutes of Health (GM33741, GM58423, GM60035, and HG02150 to D.L.H. and GM61549 to D.N.) as well as from the Deutsche Forschungsgemeinschaft (DFG Le 987/2-1 to E.A.W.).

Manuscript received September 5, 2001; Accepted for publication November 8, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ATKINSON, P. W., A. C. PINKERTON, and D. A. O'BROCHTA, 2001  Genetic transformation systems in insects. Annu. Rev. Entomol. 46:317-346[Medline].

AUGE-GOUILLOU, C., Y. BIGOT, N. POLLET, M. H. HAMELIN, and M. MEUNIERROTIVAL et al., 1995  Human and other mamm. genomes contain transposons of the mariner family. FEBS Lett. 368:541-546[Medline].

BERGHAMMER, A. J., M. KLINGLER, and E. A. WIMMER, 1999  Genetic techniques: a universal marker for transgenic insects. Nature 402:370-371[Medline].

BLACKMAN, R. K., M. M. D. KOEHLER, R. GRIMAILA, and W. M. GELBART, 1989  Identification of a fully-functional hobo transposable element and its use for germ-line transformation of Drosophila. EMBO J. 8:211-217[Medline].

BRUNET, F., T. GIRAUD, F. GODIN, and P. CAPY, 2001  Do deletions of the Mos1-like elements occur randomly in the Drosophilidae family? J. Mol. Evol. in press.

BRYAN, G., D. GARZA, and D. L. HARTL, 1990  Insertion and excision of the transposable element mariner in Drosophila. Genetics 125:103-114[Abstract].

CAPY, P., K. MARUYAMA, J. R. DAVID, and D. L. HARTL, 1991  Insertion sites of the transposable element mariner are fixed in the genome of Drosophila sechellia. J. Mol. Evol. 33:450-456[Medline].

CHIANG, S. L. and J. J. MEKALANOS, 2000  Construction of a Vibrio cholerae vaccine candidate using transposon delivery and FLP recombinase-mediated excision. Infect. Immun. 68:6391-6397[Abstract/Free Full Text].

COATES, C. J., N. JASINSKIENE, L. MIYASHIRO, and A. A. JAMES, 1998  Mariner transposition and transformation of the yellow fever mosquito Aedes aegypti. Proc. Natl. Acad. Sci. USA 95:3748-3751[Abstract/Free Full Text].

FADOOL, J. M., D. L. HARTL, and J. E. DOWLING, 1998  Transposition of the mariner element from Drosophila mauritiana in zebrafish. Proc. Natl. Acad. Sci. USA 95:5182-5186[Abstract/Free Full Text].

GARCIA-FERNÀNDEZ, J., G. MARFANY, J. BAGUÑÀ, and E. SALÓ, 1993  Infiltration of mariner elements. Nature 364:109[Medline].

GARCIA-FERNÀNDEZ, J., J. R. BAYASCAS-RAMÍREZ, G. MARFANY, A. M. MUÑOZ-MÁRMOL, and A. CASALI et al., 1995  High copy number of highly similar mariner-like transposons in planarian (Platyhelminthe): evidence for a trans-phyla horizontal transfer. Mol. Biol. Evol. 12:421-431[Abstract].

GARZA, D., M. MEDHORA, A. KOGA, and D. L. HARTL, 1991  Introduction of the transposable element mariner into the germline of Drosophila melanogaster. Genetics 128:303-310[Abstract].

GUEIROS-FILHO, F. J. and S. M. BEVERLEY, 1997  Trans-kingdom transposition: mobilization of the Drosophila element mariner within the protozoan Leishmania. Science 276:1716-1719[Abstract/Free Full Text].

GUO, B. P. and J. J. MEKALANOS, 2001  Helicobacter pylori mutagenesis by mariner in vitro transposition. FEMS Immunol. Med. Microbiol. 30:87-93[Medline].

HANDLER, A. M., 2001  A current perspective on insect gene transformation. Insect Biochem. Mol. Biol. 31:111-128[Medline].

HANDLER, A. M., S. D. MCCOMBS, M. J. FRASER, and S. H. SAUL, 1998  The lepidopteran transposon vector, piggybac, mediates germ-line transformation in the mediterranean fruit fly. Proc. Natl. Acad. Sci. USA 95:7520-7525[Abstract/Free Full Text].

HARTL, D. L., 2001  Discovery of the transposable element mariner. Genetics 157:471-476[Free Full Text].

HARTL, D. L., A. R. LOHE, and E. R. LOZOVSKAYA, 1997  Modern thoughts on an ancyent marinere: function, evolution, regulation. Annu. Rev. Genet. 31:337-358[Medline].

HORN, C. and E. A. WIMMER, 2000  A versatile vector set for animal transgenesis. Dev. Genes Evol. 210:630-637[Medline].

HORN, C., B. JAUNICH, and E. A. WIMMER, 2000  Highly sensitive, fluorescent transformation marker for Drosophila transgenesis. Dev. Genes Evol. 210:623-629[Medline].

IVICS, Z., P. B. HACKETT, R. H. PLASTERK, and Z. IZSVÀK, 1997  Molecular reconstruction of Sleeping Beauty, a Tc1-like transposon from fish, and its transposition in human cells. Cell 91:501-510[Medline].

JACOBSON, J. W., M. M. MEDHORA, and D. L. HARTL, 1986  Molecular structure of a somatically unstable transposable element in Drosophila. Proc. Natl. Acad. Sci. USA 83:8684-8688[Abstract/Free Full Text].

JARVIK, T. and K. G. LARK, 1998  Characterization of Soymar1, a mariner element in soybean. Genetics 149:1569-1574[Abstract/Free Full Text].

LIDHOLM, D.-A., G. H. GUDMUNDSSON, and H. G. BOMAN, 1991  A highly repetitive, mariner-like element in the genome of Hyalophora cecropia. J. Biol. Chem. 266:11518-11521[Abstract/Free Full Text].

LIDHOLM, D.-A., A. R. LOHE, and D. L. HARTL, 1993  The transposable element mariner mediates germline transformation in Drosophila melanogaster. Genetics 134:859-868[Abstract].

LOBO, N., X. LI, and M. J. FRASER, 1999  Transposition of the piggyBac element in embryos of Drosophila melanogaster, Aedes aegypti and Trichoplusia ni. Mol. Gen. Genet. 261:803-810[Medline].

LOHE, A. R. and D. L. HARTL, 1996a  Autoregulation of mariner transposase activity by overproduction and dominant-negative complementation. Mol. Biol. Evol. 13:549-555[Abstract].

LOHE, A. R. and D. L. HARTL, 1996b  Germline transformation of Drosophila virilis with the transposable element mariner. Genetics 143:365-374[Abstract].

LOHE, A. R. and D. L. HARTL, 1996c  Reduced germline mobility of a mariner vector containing exogenous DNA: Effect of size or site? Genetics 143:1299-1306[Abstract].

LOHE, A. R. and D. L. HARTL, 2002  Efficient mobilization of mariner in vivo requires multiple internal sequences. Genetics 160:519-526[Abstract/Free Full Text].

LOHE, A. R., D.-A. LIDHOLM, and D. L. HARTL, 1995  Genotypic effects, maternal effects and grand-maternal effects of immobilized derivatives of the transposable element mariner. Genetics 140:183-192[Abstract].

LOHE, A. R., D. DE AGUIAR, and D. L. HARTL, 1997  Mutations in the mariner transposase: the "D,D(35)E" consensus sequence is nonfunctional. Proc. Natl. Acad. Sci. USA 94:1293-1297[Abstract/Free Full Text].

LOHE, A. R., C. TIMMONS, I. BEERMAN, E. R. LOZOVSKAYA, and D. L. HARTL, 2000  Self-inflicted wounds, template-directed gap repair, and a recombination hotspot: effects of the mariner transposase. Genetics 154:647-656[Abstract/Free Full Text].

LOUKERIS, T. G., I. LIVADARAS, B. ARCÀ, S. ZABALOU, and C. SAVAKIS, 1995  Gene transfer into the Medfly, Ceratitis capitata, with a Drosophila hydei transposable element. Science 270:2002-2005[Abstract/Free Full Text].

LOZOVSKAYA, E. R., D. I. NURMINSKY, D. L. HARTL, and D. T. SULLIVAN, 1996  Germline tranformation of Drosophila virilis mediated by the transposable element hobo. Genetics 142:173-177[Abstract].

MARUYAMA, K. and D. L. HARTL, 1991a  Evidence for interspecific transfer of the transposable element mariner between Drosophila and Zaprionus. J. Mol. Evol. 33:514-524[Medline].

MARUYAMA, K. and D. L. HARTL, 1991b  Evolution of the transposable element mariner in Drosophila species. Genetics 128:319-329[Abstract].

MARUYAMA, K., K. D. SCHOOR, and D. L. HARTL, 1991  Identification of nucleotide substitutions necessary for trans-activation of mariner transposable elements in Drosophila: analysis of naturally occurring elements. Genetics 128:777-784[Abstract].

MEDHORA, M., K. MARUYAMA, and D. L. HARTL, 1991  Molecular and functional analysis of the mariner mutator element Mos1 in Drosophila. Genetics 128:311-318[Abstract].

MEDHORA, M. M., A. H. MACPEEK, and D. L. HARTL, 1988  Excision of the Drosophila transposable element mariner: identification and characterization of the Mos factor. EMBO J. 7:2185-2189[Medline].

MOREIRA, L. A., M. J. EDWARDS, F. ADHAMI, N. JASINSKIENE, and A. A. JAMES et al., 2000  Robust gut-specific gene expression in transgenic Aedes aegypti mosquitoes. Proc. Natl. Acad. Sci. USA 97:10895-10898[Abstract/Free Full Text].

MORGAN, G. T., 1995  Identification in the human genome of mobile elements spread by DNA-mediated transposition. J. Mol. Biol. 254:1-5[Medline].

NURMINSKY, D. I., E. V. BENEVOLENSKAYA, M. V. NURMINSKAYA, Y. Y. SHEVELYOV, and D. L. HARTL et al., 1998  Cytoplasmic dynein intermediate chain isoforms with different targeting properties created by tissue-specific alternative splicing. Mol. Cell. Biol. 18:6816-6825[Abstract/Free Full Text].

O'BROCHTA, D. A., W. D. WARREN, K. J. SAVILLE, and P. W. ATKINSON, 1996  Hermes, a functional non-drosophilid insect gene vector from Musca domestica. Genetics 142:907-914[Abstract].

OOSUMI, T., W. R. BELKNAP, and B. GARLICK, 1995  Mariner transposons in humans. Nature 378:672[Medline].

PATTON, J. S., X. V. GOMES, and P. K. GEYER, 1992  Position-independent germline transformation in Drosophila using a cuticle pigmentation gene as a selectable marker. Nucleic Acids Res. 20:5859-5860[Free Full Text].

PELICIC, V., S. MORELLE, D. LAMPE, and X. NASSIF, 2000  Mutagenesis of Neisseria meningitidis by in vitro transposition of Himar1. J. Bacteriol. 182:5391-5539[Abstract/Free Full Text].

PLASTERK, R. H. A., Z. IZSVAK, and Z. IVICS, 1999  Resident aliens: the Tc1/mariner superfamily of transposable elements. Trends Genet. 15:326-332[Medline].

ROBERTSON, H. M., 1993  The mariner transposable element is widespread in insects. Nature 362:241-245[Medline].

ROBERTSON, H. M., 1995  The Tc1-mariner superfamily of transposons in animals. J. Insect Physiol. 41:99-105.

ROBERTSON, H. M. and E. G. MACLEOD, 1993  Five major subfamilies of mariner transposable elements in insects, including the Mediterranean fruit fly, and related arthropods. Insect Mol. Biol. 2:125-139[Medline].

ROBERTSON, H. M., K. L. ZUMPANO, A. R. LOHE, and D. L. HARTL, 1996  Reconstructing the ancient mariners of humans. Nat. Genet. 12:360-361[Medline].

RUBIN, E. J., B. J. AKERLEY, V. N. NOVIK, D. J. LAMPE, and R. N. HUSSON et al., 1999  In vivo transposition of mariner-based elements in enteric bacteria and mycobacteria. Proc. Natl. Acad. Sci. USA 96:1645-1650[Abstract/Free Full Text].

RUBIN, G. M. and A. C. SPRADLING, 1982  Genetic transformation of Drosophila with transposable element vectors. Science 218:348-353[Abstract/Free Full Text].

SEDENSKY, M. M., S. J. HUDSON, B. EVERSON, and P. G. MORGAN, 1994  Identification of a mariner-like repetitive sequence in C. elegans. Nucleic Acids Res. 22:1719-1723[Abstract/Free Full Text].

SHERMAN, A., A. DAWSON, C. MATHER, H. GILHOOLEY, and Y. LI et al., 1998  Transposition of the Drosophila element mariner into the chicken germ line. Nature Biotech. 16:1050-1053[Medline].

SMIT, A. F. A. and A. D. RIGGS, 1996  Tiggers and other DNA transposon fossils in the human genome. Proc. Natl. Acad. Sci. USA 93:1443-1448[Abstract/Free Full Text].

THUMMEL, C. and V. PIRROTTA, 1992  New pCaSpeR P-element vectors. Dros. Inf. Serv. 71:150.

TOSI, L. R. O. and S. M. BEVERLEY, 2000  Cis and trans factors affecting Mos1 mariner evolution and transposition in vitro, and its potential for functional genomics. Nucleic Acids Res. 28:784-790[Abstract/Free Full Text].

WANG, W., L. SWEVERS, and K. IATROU, 2000  Mariner (Mos1) transposase and genomic integration of foreign gene sequences in Bombyx mori cells. Insect Mol. Biol. 9:145-155[Medline].

YOSHIYAMA, M., H. HONDA, and K. KIMURA, 2000  Successful transformation of the housefly, Musca domestica (Diptera: Muscidae) with the transposable element, mariner. Appl. Entomol. Zool. 35:321-325.

This article has been cited by other articles:

				K. J. T. Venken and H. J. Bellen Transgenesis upgrades for Drosophila melanogaster Development, October 15, 2007; 134(20): 3571 - 3584. [Abstract] [Full Text] [PDF]

				D. C. Williams, T. Boulin, A.-F. Ruaud, E. M. Jorgensen, and J.-L. Bessereau Characterization of Mos1-Mediated Mutagenesis in Caenorhabditis elegans: A Method for the Rapid Identification of Mutated Genes Genetics, March 1, 2005; 169(3): 1779 - 1785. [Abstract] [Full Text] [PDF]