JOURNAL OF BACTERIOLOGY, Mar. 1997, p. 2047–2052 0021-9193/97/$04.0010 Copyright q 1997, American Society for Microbiology
Vol. 179, No. 6
Transposition without Transposase: a Spontaneous Mutation in Bacteria CHAD A. RAPPLEYE†
AND
JOHN R. ROTH*
Department of Biology, University of Utah, Salt Lake City, Utah 84112 Received 2 October 1996/Accepted 15 January 1997
Transposition mutations are typically associated with the activities of transposable elements such as transposons and insertion sequences, whose mobility is dependent upon transposase enzymes that catalyze exchanges between element ends and target sites. We describe a single transposition event in which a block of donor sequence is inserted at a target site without the involvement of any known transposase or the ends of any known transposable element. We propose that this is a new type of spontaneous mutation which may be difficult to detect in standard mutant hunts but may be of evolutionary importance. Most spontaneous mutations can be sorted into four categories: base substitution, base addition or removal, and chromosomal rearrangements (deletions, duplications, inversions or transpositions). Rearrangements can arise by recombination between repeated sequences or can be catalyzed by transposases which act on conserved sequences at element ends. At low frequency, deletions and duplications can form without recombination between extensive repeats or transposases. Known bacterial transposition mutations are generated only by the activities of selfish elements. We describe here a spontaneous transposition event in Salmonella typhimurium that appears to have occurred without involvement of a transposable element or transposase. The mutation described moves a contiguous block of sequence (1.9 kb) and inserts it at a new site with no loss or duplication of the target sequence. While the system for detecting this mutant involved use of the transposable elements Tn10d(Tc) and Mud-lac, the mutational event itself occurred in strains devoid of any known transposase. Furthermore, the novel sequence junctions are not associated with the ends of any transposable element. Thus, the event may constitute a new type of spontaneous mutation. Rare events of this type may be difficult to detect without a specific selection system such as that used here. Spontaneous transposition events may have contributed to the evolution of hybrid proteins and multicistronic operons.
kanamycin, 50 mg/ml. Solid medium was prepared by addition of agar (1.5%; Difco) to NB or minimal medium. Genetic techniques. Transductional crosses were mediated by the high-frequency generalized transducing phage mutant P22 HT105/1 int-201 (19). Transductants were single colony isolated and made phage free by streaking on nonselective green indicator plates (4). Cross streaking to check phage sensitivity was done with the P22 clear plaque mutant H5. Selection of tetracycline-dependent Lac1 (LacTD) mutants. Mutants were selected whose lac operon was expressed from a Tn10 promoter regulated by tetracycline. The parent strain (TT17299) has closely linked Tn10d(Tc) and MudJ elements, both of which are within the ethanolamine utilization (eut) operon. The Tn10d(Tc) insertion (eutA) is promoter proximal to the MudJ insertion (eutR) and prevents expression of the lac operon present in the Mud element. Selection for Lac1 mutants was carried out by plating 0.1 ml (approximately 108 cells) from independent, saturated NB cultures onto minimal NCE– lactose–X-Gal medium containing tetracycline (2 mg/ml). Mutant Lac1 colonies arose on the plates after 2 days at 378C. Additional mutants appeared each day up to day 5. These Lac1 colonies were patched to solid NB medium and replica printed to determine the dependence of their Lac1 phenotype on the presence of tetracycline. Testing ability of the MudJ element to transpose. The MudJ element of a Lac1 revertant (TT18818) was moved to new sites in the chromosome by transposition from a P22-transduced fragment. Since the MudJ element lacks the MuA and B gene products required for transposition, these functions were provided in trans from plasmid pLP103-6-3 (from P. Van de Putte) which was present in the recipient strain (TT8353). To distinguish Kanr transposition transductants (Eut1) from recombinants that inherit the donor insertion (Eut2), the transduction cross was done on MacConkey plates containing 1% ethanolamine and 100 nM cyano-B12 and kanamycin. On this medium, Eut1 clones form red colonies and Eut2 colonies are white. Enzyme assays. Strains to be assayed were diluted 50-fold from overnight E glucose cultures into E glucose medium with or without tetracycline (2 mg/ml) and grown at 378C to mid-log phase (100 Klett units). Activity of b-galactosidase was determined as described by Miller (12) with CHCl3 (50 ml/ml) and sodium dodecyl sulfate (0.05%) to permeabilize cells.
MATERIALS AND METHODS Bacterial strains. All strains are derivatives of S. typhimurium LT2 (Table 1). The Tn10d(Tc) element is a transposition-defective derivative of transposon Tn10 lacking the transposase gene and the internal ends of IS10 (24). The MudJ element is a transposition-defective derivative of phage Mu described by Castilho et al. (3). Insertions of these elements in the eut operon have been described previously (18, 20). Media. Complex medium was nutrient broth (NB) (0.8%; Difco Laboratories) supplemented with NaCl (0.5%). Minimal medium was Vogel and Bonner E medium (5) with added glucose (0.2%) as a carbon and energy source. Minimal E medium lacking citrate (NCE) (1) supplemented with 0.2% lactose was used in selecting mutants able to grow on lactose. The chromogenic b-galactosidase substrate 5-bromo-4-chloro-3-indoyl-b-D-galactopyranoside (X-Gal) was added (final concentration, 20 mg/ml) to solid NCE lactose medium to help visualize Lac1 colonies. Antibiotics were added at the following concentrations: tetracycline, 20 mg/ml (in complex medium) or 2 mg/ml (in minimal medium), and
RESULTS TD
Selection of Lac mutants. The experimental system (Fig. 1) employs a strain (TT17299) with insertions of Tn10d(Tc) and of MudJ elements in the eut operon, which encodes genes for the utilization of ethanolamine as a carbon and/or nitrogen source (18). Within the Tn10d(Tc) element are two genes, tetA and tetR, which encode tetracycline resistance and the transcription repressor protein, respectively (reviewed in reference 9). Divergent transcription of both tet genes is induced by tetracycline. Most transcripts are terminated either within the element or in the flanking sequences close to the element (unpublished results). However, a small fraction of the transcripts extend out of the Tn10d(Tc) element into adjacent genes (22). The MudJ element contains a promoterless lacZ reporter gene that is not expressed from the eut promoter due to the polar Tn10d(Tc) insertion located 2.5 kb upstream of
* Corresponding author. Mailing address: Department of Biology, University of Utah, Salt Lake City, UT 84112. Phone: (801) 581-3412. Fax: (801) 585-6207. E-mail:
[email protected]. † Present address: Department of Biology, University of California San Diego, San Diego, CA. 2047
2048
RAPPLEYE AND ROTH
J. BACTERIOL.
TABLE 1. Bacterial strains used Strain
Genotypea
TT8353 TT17299 TT18818 TT18820 TT18822 TT18823 TT18824 TT18825 TT18826
pLP103-6-3 (Ampr MuA1 MuB1) eutA208::Tn10d(Tc) eutR156::MudJ eutA208::Tn10d(Tc)* eutR156::MudJ* eutR156::MudJ eutR156::MudJ* eutA208::Tn10d(Tc)* eutR156::MudJ eutA208::Tn10d(Tc) eutR156::MudJ* hisG10175::Tn10d(Tc) eutR156::MudJ* eutA208::Tn10d(Tc) his-10179::MudJ*
a Asterisk denotes an element derived from the original mutant strain (TT18818). Alleles without the asterisk were not subjected to selection.
the MudJ insertion site. Both transposable elements have a deletion of their respective transposase genes. Thus, both elements are defective for transposition and cannot revert to produce functional transposase. The Tn10d(Tc) element is located in the eutA gene and the MudJ element is inserted in the distal eutR gene (Fig. 1). Transcription levels from Tn10d(Tc) and from the weak promoter (PII) in the intervening eut region are insufficient to confer a Lac1 phenotype on NCE-lactose medium with or without tetracycline (data not shown). This strain was constructed to allow positive selection of mutations within the Tn10 element that allow transcription to proceed out of the element and into adjacent sequence, resulting in a tetracycline-inducible Lac1 (LacTD) phenotype. Such mutants were recovered in other experiments and will be described elsewhere (16). The transposition mutation described here was encountered during these experiments; it was selected as described below. Starting with the parent strain (TT17299), Lac1 revertants were selected by plating cells on NCE-lactose medium with added tetracycline. The frequency of Lac1 revertants was approximately 1027 per cell plated. These mutants were screened to identify the rare types whose Lac1 phenotype depended on tetracycline-inducible transcripts originating from the Tn10 promoters; such clones showed a LacTD phenotype. Events resulting in the LacTD phenotype might include (i) mutations that relieve transcription termination inside the Tn10d(Tc) element, (ii) deletions fusing the Tn10 tetR transcript and the lacZ gene without impairing production of a functional TetR protein, or (iii) transposition of a fragment containing the regulatable tet promoters into a site upstream of the lacZ gene. Three independent LacTD mutants were recovered by screening 109 cells. Two of the three mutants showed 100% linkage of the Tetr and Kanr markers in transduction crosses (data not shown); these properties are characteristic of large deletions that remove the intervening region. The remaining mutant, TT18818, showed normal linkage between the Tn10d(Tc) and MudJ markers; this mutant is described below.
The tetracycline dependence of the Lac1 phenotype suggested that lacZ transcription initiates at the tet promoter of Tn10d(Tc). The original parental strain (TT17299) expressed b-galactosidase at very low levels even in the presence of the inducer, tetracycline, indicating that efficient transcription termination occurs at some point upstream of the lacZ reporter gene. In the parent strain, tetracycline caused only a twofold increase in lacZ transcription to an absolute b-galactosidase level of 7 U (Table 2). In the LacTD mutant, uninduced transcription did not differ significantly from that of the parent, but addition of tetracycline caused a 100-fold increase in lacZ transcription (Table 2). This high level of expression was observed consistently in all strains containing the mutation. The LacTD phenotype is due to alteration of the MudJ element. Mutations within the Tn10d(Tc) element that confer the LacTD phenotype should show 100% linkage to the Tetr phenotype of Tn10d(Tc). This possibility was eliminated by showing that the LacTD phenotype and the Tetr phenotype of the Tn10d(Tc) element were separable in transductional crosses (data not shown). Furthermore, when the Tn10d(Tc) element from the LacTD mutant [designated Tn10d(Tc)*] was moved into a new eutR::MudJ background, b-galactosidase production was essentially the same as the Lac2 parent (Table 2). Thus, the tetracycline-dependent phenotype is not caused by a mutation within the Tn10d(Tc) element. Movement of the MudJ element from the LacTD mutant (designated MudJ*) into a clean eutA::Tn10d(Tc) background generated a recombinant having the full mutant phenotype. Nearly identical levels of b-galactosidase expression were seen in this reconstructed double mutant and its LacTD parent (Table 2). Linkage analysis further showed that the LacTD phenotype was 100% linked to the Kanr phenotype of the eutR:: MudJ* insertion, suggesting that the LacTD phenotype is caused by a mutation within the MudJ* element. When the MudJ* element was moved into a background lacking a Tn10 element, the resulting strain was sensitive to growth inhibition by tetracycline and thus could not be directly tested for tetracycline induction of the lacZ gene (Table 2). Introduction of an unlinked Tn10d(Tc) element restored the tetracycline resistance and this strain showed the LacTD phenotype (Table 2). In addition to providing antibiotic resistance, the unlinked Tn10d(Tc) element might also be required for production of the TetR repressor protein since the Lac phenotype is regulated by tetracycline. To test this possibility, chlortetracycline, a nontoxic analog of tetracycline, was used in strains lacking Tn10d(Tc) elements. Although nontoxic, chlortetracycline still interacts with the TetR repressor to allow transcription from the tet promoters (2). Addition of chlortetracycline reproduced the tetracycline-dependent phenotype even in strains lacking a Tn10d(Tc) element (Table 2). From these results we conclude: (i) MudJ* contains a promoter element that drives lacZ expression and responds to induction
FIG. 1. Experimental system used in detecting the transposition event. A portion of the trpDCBA sequence (indicated by trp9CBA9) precedes the lac genes within the MudJ element.
VOL. 179, 1997
TRANSPOSITION WITHOUT TRANSPOSASE
2049
TABLE 2. Induction of lacZ expression by tetracycline Strain
Relevant genotypea
TT17299 TT18818 TT18823 TT18824 TT18825 TT18826 TT18820 TT18822
eutA::Tn10d(Tc) eutR::MudJ eutA::Tn10d(Tc)* eutR::MudJ* eutA::Tn10d(Tc)* eutR::MudJ eutA::Tn10d(Tc) eutR::MudJ* hisG::Tn10d(Tc) eutR::MudJ* eutA::Tn10d(Tc) his::MudJ* eutA1 eutR::MudJ eutA1 eutR::MudJ*
b-Galactosidase activity (U) Without tetracycline
With tetracycline
With chlortetracycline
361 6 6 0.3 3 6 0.2 8 6 0.5 7 6 0.8 7 6 0.7 2 6 0.2 7 6 1.6
7 6 0.8 642 6 11 12 6 2.5 639 6 4 680 6 46 597 6 12 ndc ndc
ndb 654 6 6 nd nd nd nd 2 6 0.1 699 6 25
a
Asterisk denotes an element derived from the original mutant strain (TT18818). Alleles without the asterisk were never subjected to selection. nd, not determined. c These strains are not able to grow in the presence of tetracycline. b
by tetracycline, (ii) MudJ* lacks a functional tetA gene and cannot provide resistance to tetracycline, and (iii) MudJ* contains the tetR gene since the inducible Lac1 phenotype is seen in strains lacking any complete Tn10 element. In order to determine if the context of the MudJ insertion within the eut operon was important for production of the phenotype and if the ends of the Mud element were intact, the MudJ* element was moved to a different chromosomal site by transposition. Transposition moves only internal MudJ sequences located between the two Mu ends; no neighboring eut material should transpose. A his::MudJ* transposition mutant (phenotypically Eut1 His2) was isolated. When resistance to tetracycline was provided by a nonadjacent Tn10d(Tc) ele-
ment, the LacTD phenotype again was observed (Table 2). Thus, all determinants of the LacTD phenotype are contained within the MudJ* element. The MudJ* element contains a transposed, internal fragment of Tn10d(Tc). Using PCR primed by oligonucleotides specific for MudJ sequences and for Tn10d(Tc) sequences we determined that the MudJ* element includes an internal fragment of the Tn10d(Tc) element inserted within the trpB gene (present within the MudJ element and part of an operon including the element’s lac gene). The inserted fragment includes the tetR gene, the two tet promoters, and a portion of the tetA gene (Fig. 2). The orientation of the transposed fragment within the MudJ element was inverted with respect to
FIG. 2. Diagram showing the transposed fragment boundaries and joinpoint sequences (all sequences written 59339). Underlined bases indicate patches of sequence similarity.
2050
RAPPLEYE AND ROTH
J. BACTERIOL.
FIG. 3. Model proposed for formation of the transposition mutation. Sequences “a” to “f” are the same as those given in Fig. 2. Sequence regions indicated with primes represent the complementary sequences. Panel A is a general model. Panel B is a variant model that uses observed sequence similarities to prime replication of the transposed region from both ends. Panel C generates the mutation by replication of the transposed region primed by one broken end of the target sequence.
that of the same sequences in the original Tn10d(Tc) element so that the tetA promoter must initiate transcription of the lacZ gene of MudJ*. In order to determine the extent of this transposed Tn10d(Tc) fragment, we sequenced the end points of the transposed fragment (Fig. 2). Sequencing confirmed the PCR data and showed that only part of the tetA gene was present, supporting the Tets phenotype of strains containing only MudJ*; the transposed copy of the tetA gene lacked 177 bp and is likely to produce a nonfunctional tetracycline resistance protein. Sequencing of the tetR end of the fragment showed that a complete tetR gene was present, accounting for the regulation of the tet promoters by chlortetracycline. Analysis of the target site sequences showed that the transposed Tn10d(Tc) fragment was inserted within the trpB gene of the MudJ element. Unlike most transposition mechanisms which lead to duplication of target site sequences (7), no alteration of the trpB sequence was observed other than precise insertion of the Tn10d(Tc) fragment. The only clues to the mechanism responsible for this transposition event are the minimal sequence similarities between the ends of the donor segment (b and c in Fig. 2) and the region near the target site (e and f). These similarities are indicated by underlines (single, double or triple) in Fig. 2. Proposals for how the single and double underlined sequences might contribute to this transposition are discussed below. DISCUSSION A spontaneous transposition mutation is described that occurred without transposase enzymes or the ends of any transposable element. The described transposition mutation lacks a duplication of target site material. We believe this type of mutation has not been previously observed among bacterial mutations because it is rare and detection might require a special selection scheme. Insertion of the transposed block of DNA sequence would likely cause a null phenotype of the target gene. Standard genetic analysis of spontaneous null mutations would be unlikely to reveal this type of mutation among the much more common deletions and point mutations. In crosses, the insertion might behave as a point mutation with a
strongly polar effect on expression of downstream cistrons. We were able to detect this event because the transposed fragment contained a regulated promoter which placed a reporter gene near the target site under a specific foreign control system. In the event we describe here, transposition occurred without apparent loss of the sequence from the donor site; however, if the donor locus were on a sister chromosome, this loss would not be detected in the recovered mutant. While frequency cannot be determined from a single occurrence, the mutation described was encountered in an experiment in which 109 cells were screened. In hopes of finding more such mutations, we screened an additional 1010 cells by the same procedure without recovering another example. Thus, the event appears to be quite rare, and convenient isolation of more such mutations will require a stronger selection regimen and/or use of an appropriate mutagenic treatment. The recovered transposition mutation appears replicative in that the mutant possesses two copies of the sequence that transposed, one at the new site and one at the original position. The presence of two copies cannot be taken as evidence of replication because sister-strand recombination could reassort the donor and target regions, placing the transposition near an uninvolved copy of the donor region. Similarly, the donor and target sequences participating in the event could have been on different sister chromosomes. While both conservative and replicative transposition models can be entertained, the several sequence similarities near the junction points of the rearrangement are suggestive of some involvement of replication. We illustrate the general idea below using donor and target sites located in cis; the same events could occur between donor and target sites on different sister chromosomes. We propose that the mutation arose by repair of a doublestrand break at the target site (Fig. 3A). A simple, general model (Fig. 3A) is one in which the two 39 ends formed by a double-strand break at the target site invade on opposite sites of a distant duplex region and initiate replication across the sequence which will transpose. This model predicts that sequence similarity should exist between sequences e and a and between sequences f and d (indicated in Fig. 2 and 3). Replication primed by these 39 ends would then generate the trans-
VOL. 179, 1997
posed fragment bounded by the target site sequences (e and f); this is in essence a process of double-strand-break repair that depends on minimal sequence similarities. In the event documented here, the predicted sequence similarities were not present. Instead, a stretch of sequence similarity was found between sequences e and c (single-underlined bases in Fig. 2) and between b and f (double-underlined bases). To account for the mutation with the observed sequence repeats, a variant of this model was developed in which the target site (e to f) suffers a double-strand break by staggered single-strand cuts as indicated in Fig. 3B. The generated 39 ends can invade the duplex at donor sequence junctions and prime replication of the transposing material. While this scenario makes use of the available sequence similarities, it does not quite generate the actual product observed. In particular, this model does not generate the proper e to b junction sequence. A somewhat better variant of the model employs the same sequence repeats in a slightly different way (Fig. 3C). In this model, a staggered double-strand break in the target sequence generates two DNA ends with 59 overhangs. Fill-in of the overhangs leaves flush double-strand ends (e and e to f), both with the TGGTT sequence whose invasion of the donor region (c to d) initiates replication across the transposed sequence. The same TGGTT sequence at the opposite end of the break can contribute to formation of the e to b junction. The short sequence similarity between b and f may help to juxtapose molecule ends. Such “fill-in” models for nonhomologous DNA end joining have been proposed for joining blunt and 59 protruding single-strand termini observed in Xenopus extract systems (23). Alternatively, the TGGTT sequence of the invading e to f region can be generated by strand displacement synthesis initiated at a nick (17). Both forms of the model have a problem at the c to f junction in that primer pairing (right side of Fig. 3B and C) leaves a mispaired T residue in both primer and template strands. The proper product requires two bases (TA/ AT) at this position; this cannot be achieved by standard mismatch repair. Illegitimate (nonhomologous) recombination may be an aspect of DNA repair, particularly in eukaryotes. Many studies have analyzed mutations of the APRT locus in CHO cells following treatment with agents thought to induce chromosome breaks; these mutagens have included topoisomerase inhibitors (8), restriction enzymes (15), and gamma radiation (11). Whenever DNA breaks were induced, the set of mutants isolated have included large insertions (.100 bp) of sequences of unknown origin. Examination of a set of 120 spontaneous APRT mutations identified only one 285-bp insertion of foreign DNA (14). As with our transposition mutation, the insertion mutations at the APRT locus showed no terminal repeats or duplication of target sites as would have been expected for a transposable element. A significant correlation exists between illegitimate recombination sites and topoisomerase cleavage sites, which may indicate that double-strand breaks generated by these enzymes are important in nonhomologous recombination events. The frequency of deletion, duplication, and insertion mutations was increased by treatment of eukaryotic cells with drugs that permit topoisomerase II to cleave but not religate DNA (8). Chromosome translocation joinpoints coincide with topoisomerase cleavage sites in leukemia cells arising following treatment with topoisomerase-targeted drugs (6). Similarly, illegitimate recombination events in lambda prophage induction correlate with DNA gyrase cleavage sites following treatment with oxolinic acid (21). The involvement of topoisomerase II has led to a topoisomerase subunit exchange model to explain the joining
TRANSPOSITION WITHOUT TRANSPOSASE
2051
of nonhomologous DNA molecules (10). In this model, topoisomerase II binds and transiently cleaves a DNA duplex leaving a topoisomerase monomer attached to the 59 end of each DNA duplex terminus. Subsequent dimerization with monomers attached to a second DNA molecule results in exchange of DNA strands. Such a cut-and-paste model would account for the lack of sequence similarity observed at the joinpoints of our transposition mutant. The small stretches of sequence similarity neighboring the joinpoints may have helped to juxtapose topoisomerase subunits from the donor and target DNA duplexes. The repair of DNA lesions by incorporation of ectopic sequences may be a general mechanism of double-strand-break repair that is an alternative to homologous recombination. This foreign DNA could come from many sources. As was proposed in Fig. 3C, transposed sequences could be copied from information contained within intact chromosomes. Inserted DNA could also arise from the capture of free linear DNA fragments present in the cell, such as Okazaki fragments. Recently, analysis of nonhomologous double-strand-break repair in Saccharomyces cerevisiae showed about 1% of repair events incorporated cDNA fragments of transposon Ty1 messenger RNA at the DNA break site (13). Analysis of the informational content of transposed DNA in these types of mutants may also provide clues to the precise mechanism of their formation. Spontaneous transpositions of the sort described may play a general role in the evolution of genomes. As was observed in the specific event described here, natural transposition can place genes under control of alternative promoters. Transposition of an intact gene into regions downstream of a promoter could form multicistronic operons. Similarly, natural transposition events could assemble fragments of genes leading to production of multifunctional or multidomain proteins. ACKNOWLEDGMENT This work was supported in part by a grant (GM27068) from the National Institutes of Health. REFERENCES 1. Berkowitz, D., J. M. Hushon, H. J. Whitfield, J. Roth, and B. N. Ames. 1968. Procedure for identifying nonsense mutations. J. Bacteriol. 96:215–220. 2. Bochner, B. R., H. Huang, G. L. Schieven, and B. N. Ames. 1980. Positive selection for loss of tetracycline resistance. J. Bacteriol. 143:926–933. 3. Castilho, B. A., P. Olfson, and M. J. Casadaban. 1984. Plasmid insertion mutagenesis and lac gene fusion with mini-Mu bacteriophage transposons. J. Bacteriol. 158:488–495. 4. Chan, R. K., D. Botstein, T. Watanabe, and Y. Ogata. 1972. Specialized transduction of tetracycline resistance by phage P22 in Salmonella typhimurium. II. Properties of a high-frequency-transducing lysate. Virology 50: 883–898. 5. Davis, R. W., D. Botstein, and J. R. Roth. 1980. Advanced bacterial genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 6. Felix, C. A., B. J. Lange, M. R. Hosler, J. Fertala, and M. Bjornsti. 1995. Chromosome band 11q23 translocation breakpoints are DNA topoisomerase II cleavage sites. Cancer Res. 55:4287–4292. 7. Galas, D. J., and M. Chandler. 1989. Bacterial insertion sequences, p. 109– 162. In D. E. Berg and M. M. Howe (ed.), Mobile DNA. American Society for Microbiology, Washington, D.C. 8. Han, Y., M. J. F. Austin, Y. Pommier, and L. F. Povirk. 1993. Small deletion and insertion mutations induced by the topoisomerase II inhibitor teniposide in CHO cells and comparison with sites of drug-stimulated DNA cleavage in vitro. J. Mol. Biol. 229:52–66. 9. Hillen, W., and C. Berens. 1994. Mechanisms underlying expression of Tn10 encoded tetracycline resistance. Annu. Rev. Microbiol. 48:345–369. 10. Ikeda, H., K. Moriya, and T. Matsumoto. 1981. In vitro study of illegitimate recombination: involvement of DNA gyrase. Cold Spring Harbor Symp. Quant. Biol. 45:399–408. 11. Miles, C., G. Sargent, G. Phear, and M. Meuth. 1990. DNA sequence analysis of gamma radiation-induced deletions and insertions at the APRT locus of hamster cells. Mol. Carcinogenesis 3:233–242. 12. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor
2052
RAPPLEYE AND ROTH
Laboratory, Cold Spring Harbor, N.Y. 13. Moore, J. K., and J. E. Haber. 1996. Capture of retrotransposon DNA at the sites of chromosomal double-strand breaks. Nature 383:644–646. 14. Nalbantoglu, J., C. Miles, and M. Meuth. 1988. Insertion of unique and repetitive DNA fragments into the aprt locus of hamster cells. J. Mol. Biol. 200:449–460. 15. Phillips, J. W., and W. F. Morgan. 1994. Illegitimate recombination induced by DNA double-strand breaks in a mammalian chromosome. Mol. Cell. Biol. 14:5794–5803. 16. Rappleye, C. A., and J. R. Roth. Submitted for publication. 17. Ripley, L. S. 1994. Deletion and duplication sequences induced in CHO cells by teniposide (VM-26), a topoisomerase II targeting drug, can be explained by the processing of DNA nicks produced by the drug-topoisomerase interaction. Mutat. Res. 312:67–78. 18. Roof, D. M., and J. R. Roth. 1988. Ethanolamine utilization in Salmonella typhimurium. J. Bacteriol. 170:3855–3863. 19. Schmieger, H. 1971. A method for detection of phage mutants with altered
J. BACTERIOL. transducing ability. Mol. Gen. Genet. 110:378–381. 20. Sheppard, D. E., and J. R. Roth. 1994. A rationale for autoinduction of a transcriptional activator: ethanolamine ammonia-lyase (EutBC) and the operon activator (EutR) compete for adenosyl-cobalamin in Salmonella typhimurium. J. Bacteriol. 176:1287–1296. 21. Shimizu, H., H. Yamaguchi, and H. Ikeda. 1995. Molecular analysis of lambda-bio transducing phage produced by oxolinic acid-induced illegitimate recombination in vivo. Genetics 140:889–896. 22. Takiff, H. E., T. Baker, T. Copeland, S. Chen, and D. L. Court. 1992. Locating essential Escherichia coli genes by using mini-Tn10 transposons: the pdxJ operon. J. Bacteriol. 174:1544–1553. 23. Thode, S., A. Schafer, P. Pfeiffer, and W. Vielmetter. 1990. A novel pathway of DNA end-to-end joining. Cell 60:921–928. 24. Way, J. C., M. A. Davis, D. Morisato, D. E. Roberts, and N. Kleckner. 1984. New Tn10 derivatives for transposon mutagenesis and for construction of lacZ operon fusions by transposition. Gene 32:369–379.
