- Email: [email protected]
Isolation and characterization of IS1 circles
Gene 191 ( 1997) 183–190
Isolation and characterization of IS1 circles Yasuhiko Sekine, Naoki Eisaki, Kiyoaki Kobayashi, Eiichi Ohtsubo * Institute of Molecular and Cellular Biosciences, The University of Tokyo, Bunkyo-ku, Yayoi 1-1-1, Tokyo 113, Japan Received 16 August 1996; received in revised form 2 December 1996; accepted 14 December 1996; Received by A. Nakazawa
Abstract Transposase encoded by insertion sequence IS1 is produced from two out-of-phase reading frames by translational frameshifting that occurs in a run of adenines. An IS1 mutant with a single adenine insertion in the run of adenines eciently produces transposase, resulting in generation of miniplasmids by deletion for a region adjacent to IS1 from a plasmid carrying the IS1 mutant. Here, we found that besides miniplasmids, cells harboring the plasmid contained minicircles without the region required for replication. Cloning and DNA sequencing of the minicircles revealed that most of them were IS1 circles consisting of the entire IS1 sequence and a sequence, 5–9 bp in length, which intervenes between terminal inverted repeats, IRL and IRR, of IS1. Analysis of more IS1 circles isolated by polymerase chain reaction revealed that the intervening sequence was derived from the region flanking either IRL or IRR in the parental plasmid, suggesting that IS1 circles are generated by an excision event from the parental plasmid. The IS1 circles may be formed due to the cointegration reaction occurring within the parental plasmid carrying IS1. © 1997 Elsevier Science B.V. Keywords: Translational frameshifting; IS1 transposase; Minicircles; Cointegration; Simple insertion
1. Introduction Insertion sequence IS1 is the smallest active transposable element (768 bp) in bacteria and carries imperfect terminal inverted repeats, IRL and IRR, of about 30 bp long (Ohtsubo and Ohtsubo, 1978; Johnsrud, 1979). IS1, which appears as a simple insertion in bacterial and phage genes (Hirsch et al., 1972; Fiandt et al., 1972; Calos et al., 1978; Grindley, 1978 ), can mediate cointegration between a donor replicon carrying IS1 and a recipient replicon (Ohtsubo et al., 1980, 1981; Iida and Arber, 1980; Galas and Chandler, 1982). Simple insertion of IS1 is assumed to occur due to cleavage of the strands of the donor replicon in the strand-transfer product, an intermediate molecule of cointegration, followed by release and degradation of the donor portion (Ohtsubo et al., 1981). The cointegration reaction that occurs intramolecularly is believed to generate a deletion of a DNA segment adjacent to IS1 (Ohtsubo et al., * Corresponding author. Tel./fax: +81 3 56843269; e-mail: [email protected] Abbreviations: bp, base pair (s); IR(s), terminal inverted repeat (s); IRL and IRR, terminal inverted repeats at left and right, respectively; IS, insertion sequence; nt, nucleotide(s); R , resistant/resistance; Km, kanamycin; Cm, chloramphenicol; PCR, polymerase chain reaction. 0378-1119/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S0 3 78 - 11 19 ( 9 7 ) 00 05 7 -7
1978; Sekine and Ohtsubo, 1991; Sekino et al., 1995; Turlan and Chandler, 1995). Upon cointegration, IS1 gives rise to duplication of the target sequence, mostly 9 bp in length (Ohtsubo et al., 1980; Machida and Machida, 1987 ). IS1 encodes two open reading frames, insA and insB (Machida et al., 1982, 1984; Jakowec et al., 1988). A frameshifting event in the −1 direction from the 3∞-end region of insA to an open reading frame B∞-insB, in which B∞ is the frame extended from the 5∞ end of insB and overlaps insA in −1 phase, is involved in production of the InsA-B∞-InsB transframe protein that is IS1 transposase (Sekine and Ohtsubo, 1989; Escoubas et al., 1991). The frameshifting occurs in a run of six adenines in the overlapping region between insA and B∞-insB (Sekine et al., 1992). An IS1 mutant (IS1-31) with a single adenine insertion in the run of adenines, which results in in-frame alignment of insA and B∞-insB, eciently produces transposase without frameshifting. This mutant in a plasmid mediates cointegration at a frequency much higher than wild-type IS1 (Sekine and Ohtsubo, 1989). The mutant also mediates deletion to give rise to a large amount of miniplasmids from the plasmid carrying it (Sekine and Ohtsubo, 1991; Sekino et al., 1995 ). Here, we report that the plasmid carrying IS1-31 with
184
Y. Sekine et al. / Gene 191 (1997) 183–190
a single adenine insertion generates miniplasmids as well as characteristic IS1 circles, consisting of the entire IS1 sequence and a short intervening sequence between IRL and IRR of IS1. The IS1 circles identified here may correspond to an excised circular copy of the transposon with two end regions of IS1, which was observed in the cells where transposase was produced eciently ( Turlan and Chandler, 1995). We present models for generation of the IS1 circles.
2. Materials and methods 2.1. Bacterial strains and plasmids Bacterial strains used were Escherichia coli K12 derivatives, YK1100 ( Wada et al., 1988 ), MV1184 (Vieira and Messing, 1987 ), MC4100 (Silhavy, 1984 ), and NM554 (Raleigh et al., 1988 ). Plasmid pSEK131 used was a pUC119-derivative carrying IS1-31, an IS1 mutant with a single adenine insertion in the run of adenines (Sekine et al., 1992). Plasmid pSEK117 is a pUC119-derivative carrying wild type IS1 (Sekine et al., 1992). Plasmid pSEK131TI, which is a type I miniplasmid derived from pSEK131, is deleted for the region (973 bp) extending from the end of IRL to a site located upstream of the bla gene (Sekine et al., 1992). Plasmid pHSG398 ( Takeshita et al., 1987) is a pUC18-type plasmid used for cloning of minicircles. Plasmid DNA was prepared from bacterial cells grown in L-rich broth. Small-scale preparation of plasmid DNA was performed as described by He et al. (1990). The alkaline lysis method (Sambrook et al., 1989) was used to prepare plasmid DNA for cloning and nucleotide sequencing. 2.2. Media Culture media used were L broth, L-rich broth and w-medium ( Yoshioka et al., 1987). w-Medium was used for transformation of plasmid DNA. L-agar plates contained 1.5% (w/v) agar (Eiken) in L broth. Antibiotics were added in L-agar plates, when necessary, at the concentration of 100 mg ampicillin ( Wako)/ml, 30 mg kanamycin (Sigma)/ml, or 30 mg chloramphenicol (Sigma)/ml. 2.3. Nucleotide sequencing Nucleotide sequences were determined by the dideoxynucleotide method (Sanger et al., 1977; Messing, 1982) using the Sequenase DNA sequencing kit ( U.S. Biochemical Corp.). Oligonucleotides corresponding to the sequence at nt 70 to 51 of the lower strand (named IS1-L) and to the sequence at nt 704 to 723 of the upper strand (named IS1-R) of IS1 with coordinates 1–768
(Ohtsubo and Ohtsubo, 1978) were used as primer for sequencing. These oligonucleotides were synthesized using a DNA synthesizer model 392 (Applied Biosystems). The DNA chains were labeled with [a-32P]dCTP (15 TBq/mmol, Amersham) and separated on 8% polyacrylamide gels containing 7 M urea. 2.4. Cloning of IS1 circles IS1 circles were cloned as follows: Total closed circular DNA was prepared from each of two clones of the YK1100 cells harboring pSEK131 or each of two clones of the MV1184 cells harboring pSEK131. The DNA was digested with PstI (Takara), which cleaves pSEK131 at one site within IS1, and subjected to electrophoresis in a 5% polyacrylamide gel. The DNA (approximately 0.8 kb) eluted from the gel was inserted into the PstI site of pHSG398. The DNA sample was introduced into NM554 by transformation and the CmR transformants were selected. Plasmids in several CmR transformants originated from the DNA sample from a cell culture were prepared and analyzed for their structures by sequencing. Critical regions in the IS1 circles containing an intervening sequence were cloned as follows: The total closed circular DNA prepared from each of three clones of the MV1184/pSEK131 cells, each of four clones of the MC4100/pSEK131TI cells, or each of four clones of the NM554/pSEK131TI cells was digested with PstI. Using the DNA preparation (0.5 ng) as template and a pair of primers IS1-L and IS1-R (50 pmol each), polymerase chain reaction ( PCR) was carried out in 50 ml of Cetus buer containing 2.5 units of AmpliTaq DNA polymerase ( Perkin Elmer Cetus). Incubations at 93°C (2 min), 50°C (2 min), and 72°C (3 min) were repeated 28 times in an automated thermal cycler. The DNA fragment amplified was inserted into the SmaI site of pUC119, and the plasmid DNA sample was introduced into NM554 by transformation.
3. Results and discussion 3.1. Identification of IS1 circles in cells harboring pSEK131 Plasmid pSEK131 (3.9 kb in length) is a derivative of pUC119 carrying IS1-31. Total plasmid DNA molecules, which were prepared from cells of an E. coli strain YK1100(recA+) harboring this plasmid, contained miniplasmids of dierent sizes, approximately 2.8 and 2.0 kb in length (Fig. 1A), while those from cells harboring pSEK117, another pUC119-derivative carrying wild type IS1, did not (Fig. 1B). These miniplasmids are supposed to correspond to those, called type I and II, previously identified and characterized (see shadowed
Y. Sekine et al. / Gene 191 (1997) 183–190
185
Fig. 1. Ethidium bromide-stained gels showing production of miniplasmids and minicircles. (A) An agarose gel (0.7%) showing total closed circular DNA prepared from the YK1100 cells harboring pSEK117 or pSEK131. ( B) An agarose gel ( 0.7%) showing the two DNA preparations digested with PstI. (C ) A polyacrylamide gel ( 5%) showing the PstI digests of total closed circular DNA prepared from the YK1100 cells harboring pSEK131. Positions of the parental plasmid DNA, miniplasmids (type I and II ) and minicircles ( type III ) are indicated. Note that minicircles are not visualized in the agarose gels because of small amounts loaded.
boxes in Fig. 2), with a deletion extending from the IRL end of IS1 (Sekino et al., 1995 ). We then prepared a large amount of total plasmid DNA from the YK1100 cells harboring pSEK131. Electrophoresis in a polyacrylamide gel revealed that the DNA sample contained not only miniplasmids but also minicircular DNA molecules (here called type III; Fig. 2 ), approximately 0.8 kb in length, which could be linearized by digestion with PstI ( Fig. 1C). The minicircular DNA molecules digested with PstI were eluted from the gel, ligated with the PstI fragment containing the kanamycin-resistance ( KmR) gene, and introduced into YK1100 by transformation. We obtained no KmR transformants, showing that the minicircles, unlike miniplasmids (Sekino et al., 1995 ), cannot replicate autonomously due to lack of the region required for DNA replication. To analyze structures of the minicircles, the minicircular DNA molecules digested with PstI were inserted into the PstI site of a chloramphenicol-resistance (CmR) plasmid pHSG398. The minicircular DNA molecules were also prepared from cells of another E. coli strain MV1184 (recA−) harboring pSEK131 and inserted into pHSG398. DNA sequencing analysis of plasmids in the CmR transformants revealed that eleven out of seventeen clones obtained contained minicircles with a structure consisting of the entire IS1-31 sequence and an intervening sequence (6–9 bp in length) between IRL and IRR of IS1 ( Figs. 2 and 3A). We call these minicircles as IS1 circles, since they were distinct from the other minicircles (see legend to Fig. 2 ). Of eleven clones of the IS1 circles,
five clones had an intervening sequence identical to that flanking either IRR or IRL of IS1 in the parental plasmid pSEK131 (see clones 7, 9, M6, M8, and M10 in Fig. 3A). Four clones had an intervening sequence which was present at a region distant from IS1 in pSEK131 (see clones 4, 6, M1, and M9 in Fig. 3A). The intervening sequences in the other two clones were, however, not found in pSEK131 (see clones 13 and 31 in Fig. 3A). There was no significant structural dierence among the IS1 circles generated in both host strains (recA+ and recA−) used here, suggesting that the RecA function is not involved in formation of the IS1 circles.
3.2. Characterization of IS1 circles To characterize the IS1 circles, we examined more IS1 circles by isolating their portions by polymerase chain reaction (PCR) using total plasmid DNA prepared from independent cultures of the MV1184 cells harboring pSEK131 as template and a pair of primers that hybridize with IRL and IRR and prime DNA synthesis toward outside of IS1. DNA sequencing analysis of the amplified fragments cloned into the SmaI site of pUC119 revealed that all the clones from each of three samples from independent cultures contained the fragments with a structure consisting of two end regions of IS1 and an intervening sequence between IRL and IRR, 6–9 bp in length (Fig. 3B). Five clones contained the intervening sequence identical to that flanking IRR of IS1, whereas
186
Y. Sekine et al. / Gene 191 (1997) 183–190
Fig. 2. Structures of miniplasmids and minicircles derived from plasmid pSEK131. pSEK131 is a pUC119-derivative carrying IS1-31 with IRL and IRR (triangles) and is shown in a linear representation. Italic letters are coordinates given to pSEK131 (1–3926) with IS1-31 ( 898–1666). The coding region for the bla gene (a thick arrow) and a primer for DNA replication, RNA II (thin arrow), are shown. ori, the replication origin of pUC119. A restriction site for PstI located within IS1 is shown. The nucleotide sequence of a region in IS1-31 containing the run of adenines with a single adenine insertion, which results in in-frame fusion between insA and B∞-insB, is shown. Shaded thick lines represent approximate structures of miniplasmids (type I and II ) and minicircles (type III ). The type I and II miniplasmids have a sequence starting from the exact end at IRL of IS1-31 to a site within a restricted region indicated at the right side of the structures of miniplasmids (Sekino et al., 1995). Thin lines indicate structures of the cloned minicircles. Numbers with and without M indicate minicircles derived from pSEK131 in MV1184 and YK1100, respectively. IS1 circles among minicircles had a unique structure with an intervening sequence which is shown in Fig. 3A. Of the other minicircles, two are those with the entire IS1 sequence and a pUC119 sequence, 35 or 59 bp in length, which is much longer than those in the IS1 circles (see clones 10 and 12). Four are those with IS1 which lacked a terminal region of either IRL or IRR at positions indicated (see clones 1, M2, M7, and M3). These minicircles are supposed to be the products of the deletion extending from one end of IS1-31 to a site near or within the other IS1-31 end region.
another five contained the intervening sequence identical to that flanking IRL of IS1, and one clone contained a 9-bp sequence identical to that duplicated at both ends of IS1 (Fig. 3B). The intervening sequence in another IS1 circle, however, was found at a region distant from IS1 in the parental plasmid pSEK131 (Fig. 3B). Among the IS1 circles identified in this or the previous section, some had an intervening sequence which is identical to the sequence flanking IRL or IRR of IS1 in the parental plasmid. This suggests that they are formed
directly from the parental plasmid by an excision event. The other IS1 circles, however, had an intervening sequence (shown by thin letters in Fig. 3 ) which is dierent from the sequence flanking IRL or IRR and from one another. Many of such intervening sequences were found in the sequence of the parental plasmid at the regions near the points of deletion to give rise to type I and II miniplasmids (see Figs. 2 and 3 ). This suggests that these IS1 circles are generated from miniplasmids present in cells harboring pSEK131.
Y. Sekine et al. / Gene 191 (1997) 183–190
187
Fig. 4. Nucleotide sequences intervening between IRL and IRR in the IS1 circles derived from a miniplasmid pSEK131TI in MC4100 or NM554. A critical portion of pSEK131TI is shown at top. Note that the 9-bp sequences ( boldface letters) flanking IS1-31 are dierent each other. For the other information, see legend to Fig. 3.
3.3. IS1 circles generated from a miniplasmid
Fig. 3. Structures of the IS1 circles with a sequence intervening between IRL and IRR. A critical portion of the parental plasmid pSEK131 containing IS1-31 with IRL and IRR ( triangles) and the flanking regions is shown at top. A 9-bp sequence ( boldface letters) is duplicated at the regions adjacent to IS1-31. The IS1 circles contain a nucleotide sequence, 6–9 bp in length, that intervenes between IRL and IRR. The IS1 circles are shown in a linear representation by cutting them at the point between an IR and the intervening sequence. The IS1 circles shown in ( A) are depicted schematically in Fig. 2. The IS1 circles shown in (B) were actually obtained as the fragments containing the intervening sequences, which were amplified by PCR and inserted into pUC119 (see section 2: Materials and methods). These IS1 circles have numbers, 1–3, which indicate clones isolated from independent cultures of the MV1184/pSEK131 cells. Alphabetic letters (a–e) are clones from the same culture of cells. The intervening sequences, which are identical to a sequence flanking an IR of IS1-31, are indicated by boldface uppercase letters. The intervening sequences indicated by the non-boldface uppercase letters are present in pSEK131 at positions as indicated in parentheses (see map of pSEK131 in Fig. 2). The intervening sequences in clones 4, M1, M9 and 1a are positioned on the side of IRL, because the IS1 circles with these sequences are supposed to be derived from miniplasmids with a deletion extending from IRL, whereas the intervening sequence in clone 6 is positioned on the side of IRR, because the IS1 circle with the sequence is supposed to be
Plasmid pSEK131TI is a type I miniplasmid derived from pSEK131 by deletion for the region ( 973 bp) extending from the end of IRL to a site located upstream of the bla gene (see Fig. 2 for the structure of pSEK131). pSEK131TI thus has an IRL-flanking sequence dierent from that in the parental plasmid (see Fig. 4), but still generates smaller miniplasmids and the IS1 circles (data not shown). To analyze the IS1 circles, we carried out PCR using total plasmid DNA prepared from either MC4100 (recA+) or NM554 (recA−) harboring pSEK131TI as template and a pair of primers that hybridize with IRL and IRR, and obtained 21 clones of the fragments containing the intervening sequences in the IS1 circles. DNA sequencing revealed that the fragments in the clones consisted of two end regions of IS1 and an intervening sequence between IRL and IRR with 5–8 bp in length (Fig. 4 ). Among them, ten clones contained the intervening sequence identical to that derived from a miniplasmid with a deletion extending from IRR. The intervening sequences shown by lowercase letters in clones 13 and 31 are not present in pSEK131. Note that positioning of the intervening sequences on the side of IRR is tentative, because the IS1 circles with these sequences are supposed to be derived from the E. coli chromosome where IS1-31 had been inserted.
188
Y. Sekine et al. / Gene 191 (1997) 183–190
flanking IRR in the parental plasmid pSEK131TI, whereas one contained the intervening sequence identical to that flanking IRL ( Fig. 4 ), supporting the previous suggestion that these are formed by an excision event from the miniplasmids. The intervening sequences in seven clones were not identical to those flanking IS1-31, but were found in pSEK131TI near the junction points of deletion almost exclusively in the type II miniplasmids ( Fig. 4 ). This suggests that these IS1 circles are generated by excision from miniplasmids present in cells harboring pSEK131TI. Three clones contained the intervening sequences, which were not found in pSEK131TI (Fig. 4 ). Such clones have been also found in the IS1 circles previously identified as described above. We assume that these are chromosomal sequences which became adjacent to IS1-31 transposed from pSEK131TI (or miniplasmids), and then were incorporated into the IS1 circles. It should be noted here that these IS1 circles were not derived from the IS1 elements already resident in the E. coli chromosome ( Umeda and Ohtsubo, 1991), since the nucleotide sequences of the intervening sequences in the IS1 circles were not the same as those flanking each IS1 element in the chromosome, and since the internal regions of several IS1 circles examined were found to contain an adenine insertion characteristic of IS1-31. Here again, there was no significant structural dierence between the IS1 circles generated in both strains (recA+ and recA−), confirming that the RecA function is not involved in formation of the IS1 circles. 3.4. Possible models for formation of the IS1 circles How are the IS1 circles formed? IS1 forms cointegrates between a donor molecule carrying IS1 and a recipient molecule. In the first step of cointegration, each strand of the donor molecule is assumed to be cut at the 3∞ end of IS1 to generate a pair of 3∞-OH termini, each of which is transferred to each strand at a target site sequence in the recipient molecule. The IS1 circles may be formed due to the cointegration reaction occurring in the donor molecules, such that one of the 3∞-OH termini is transferred to the same strand of a donor molecule at a site in the sequence flanking another end of IS1, followed by DNA synthesis initiated from the other 3∞-OH terminus to fill the gap in the opposite strand and by cleavage to remove the non-IS1 sequence (model A in Fig. 5 ). Alternatively, each of the 3∞-OH termini might be transferred to a site in the sequence flanking the other end of IS1. If the sequences flanking IS1 are dierent from each other, the intervening sequence in the IS1 circle will be a heteroduplex form (model B in Fig. 5 ). It is also possible that in the initial step of cointegration, only one strand happens to be cut at the 3∞ end of IS1, and that the 3∞-OH terminus is transferred to the sequence flanking the other IS1 end,
leading to formation of an IS1 circle via cleavage of a strand to give the 3∞ end used for repair synthesis (model C in Fig. 5). Although we have no experimental evidence to determine which model is the case, these reactions may occur in the DNA-protein complex in which both IRs are held together by transposase; in such a situation, the 3∞-OH terminus at one end generated by transposase is located near the sequence flanking another end of IS1, which is the target of its attack, to result in generation of the IS1 circles with a short intervening sequence, 5–9 bp in length. We have observed in this study that the IS1 circles in some clones isolated independently had the same intervening sequence (Figs. 3 and 4). This suggests that some sites are more preferentially used as the target for the attack of the 3∞-OH terminus to generate the IS1 circles than other sites. Turlan and Chandler (1995) have recently reported that excised circular molecules of the transposon with two end regions of IS1 are present in the cells where transposase is produced eciently. These molecules had an intervening sequence like the IS1 circles identified here. The excised circular molecules, however, had the same intervening sequence, 8 bp in length, in the region flanking IRR of IS1 (Turlan and Chandler, 1995). This suggests that only one particular site has been used as the target for the attack of the 3∞-OH terminus to result in generation of the excised transposon molecules. IS3 and IS911 are closely related elements, which are distinct from IS1, since they transpose in a nonreplicative manner and thus do not form cointegrates. These elements, however, can generate characteristic IS circles consisting of the entire IS sequence and a 3-bp sequence intervening between both IRs, when transposases are eciently produced ( Polard et al., 1992; Sekine et al., 1994). In the case of IS911, strand-transfer products depicted in model C in Fig. 5 are observed, leading to the assumption that such molecules are intermediates of the IS911 circles (Polard and Chandler, 1995). 3.5. Role of the IS circles What is the role of IS circles? We have observed that IS3 generates linear molecules with a pair of 3∞-OH termini of IS3 (Sekine et al., 1996 ). We have also observed that the IS3 circles are linearized by a staggered break introduced in the junctions between the intervening sequence and either IRL or IRR of IS3 (Aihara, K., Sekine, Y. and Ohtsubo, E., unpublished results). It has been proposed that the linear molecules are subsequently transferred to a target molecule to generate a simple insertion product like integration of retroviruses upon retroposition (Sekine et al., 1996 ). In IS1, linear molecules have not, however, found, suggesting that IS1 circles can transpose in a dierent mode from IS3 not through linearization of the circles, or that the IS1
Y. Sekine et al. / Gene 191 (1997) 183–190
189
Fig. 5. Possible models ( A–C) for formation of the IS1 circles. Two strands of IS1 are shown by thick lines. Small arrows indicate positions of a nick introduced in a DNA strand. Broken arrows indicate transfer of the 3∞-OH termini. Wavy arrows indicate displacement DNA synthesis.
circles are the end products, which are inert for transposition. Further experiments are required to elucidate the role of the IS1 circles.
Acknowledgement This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan.
4. Conclusions (1 ) E. coli cells harboring a plasmid with an IS1 mutant, which eciently produces transposase, contained minicircles lacking the region required for replication. Most of them were IS1 circles with a characteristic structure with the entire IS1 sequence and a sequence, 5–9 bp long, intervening between IRL and IRR of IS1. (2 ) Most of the intervening sequences in the IS1 circles are identical to the sequence flanking either IRL or IRR in the parental plasmid, indicating that such IS1 circles were generated by an excision event from the parental plasmid. (3 ) The IS1 circles may be formed due to the cointegration reaction occurring in a parental plasmid containing IS1.
References Calos, M.P., Johnsrud, L., Miller, J.H., 1978. DNA sequence at the integration sites of the insertion element IS1. Cell 13, 411–418. Escoubas, J.M., Pre`re, M.F., Fayet, O., Salvignol, I., Galas, D., Zerbib, D., Chandler, M., 1991. Translational control of transposition activity of the bacterial insertion sequence IS1. EMBO J. 10, 705–712. Fiandt, M., Szybalski, W., Malamy, M.H., 1972. Polar mutations in lac, gal, phage l consist of a few IS-DNA sequences inserted with either orientation. Mol. Gen. Genet. 119, 223–231. Galas, D.J., Chandler, M., 1982. Structure and stability of Tn9-mediated cointegrates. Evidence for two pathways of transposition. J. Mol. Biol. 154, 245–272. Grindley, N.D.F., 1978. IS1 insertion generates duplication of a nine base pair sequence at its target site. Cell 13, 419–426. He, M., Wilde, A., Kaderbhai, M.A., 1990. A simple single-step
190
Y. Sekine et al. / Gene 191 (1997) 183–190
procedure for small-scale preparation of Escherichia coli plasmid. Nucleic Acids Res. 18, 1660. Hirsch, H., Starlinger, P., Brachet, P., 1972. Two kinds of insertions in bacterial genes. Mol. Gen. Genet. 119, 191–206. Iida, S., Arber, W., 1980. On the role of IS1 in the formation of hybrids between the bacteriophage P1 and the R plasmid NR1. Mol. Gen. Genet. 177, 261–270. Jakowec, M., Prentki, P., Chandler, M., Galas, D., 1988. Mutational analysis of the open reading frames in the transposable element IS1. Genetics 120, 47–55. Johnsrud, L., 1979. DNA sequence of the transposable element IS1. Mol. Gen. Genet. 169, 213–218u Machida, Y., Machida, C., Ohtsubo, E., 1984. Insertion element IS1 encodes two structural genes required for its transposition. J. Mol. Biol. 177, 229–245. Machida, C., Machida, Y., 1987. Base substitutions in transposable element IS1 cause DNA duplication of various length at the target site for plasmid cointegration. EMBO J. 6, 1799–1803. Machida, Y., Machida, C., Ohtsubo, H., Ohtsubo, E., 1982. Factors determining frequency of plasmid cointegration mediated by insertion sequence IS1. Proc. Natl. Acad. Sci. USA 79, 277–281. Messing, J., 1982. New M13 vectors for cloning. Methods Enzymol. 101, 20–78. Ohtsubo, H., Ohtsubo, E., 1978. Nucleotide sequence of an insertion element, IS1. Proc. Natl. Acad. Sci. USA 75, 615–619. Ohtsubo, E., Rosenbloom, H., Schrempf, H., Goebel, W., Rosen, J., 1978. Site specific recombination involved in the generation of small plasmids. Mol. Gen. Genet. 159, 131–141. Ohtsubo, E., Zenilman, M., Ohtsubo, H., 1980. Plasmids containing insertion elements are potential transposons. Proc. Natl. Acad. Sci. USA 77, 750–754. Ohtsubo, E., Zenilman, M., Ohtsubo, H., McCormick, M., Machida, C., Machida, Y., 1981. Mechanism of insertion and cointegration mediated by IS1 and Tn3. Cold Spring Harbor Symp. Quant. Biol. 45, 283–295. Polard, P., Chandler, M., 1995. An in vivo transposase-catalyzed single-stranded DNA circularization reaction. Genes Dev. 9, 2846–2858. Polard, P., Pre`re, M.F., Fayet, O., Chandler, M., 1992. Transposaseinduced excision and circularization of the bacterial insertion sequence IS911. EMBO J. 11, 5079–5090. Raleigh, E.A., Murray, N.E., Revel, H., Blumenthal, R.M., Westaway, D., Reith, A.D., Rigby, P.W., Elhai, J., Hanahan, D., 1988. McrA and McrB restriction phenotype of some E. coli strains and implications for gene cloning. Nucleic Acids Res. 16, 1563–1574. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 1.25–1.28.
Sanger, F., Nicklen, S., Coulson, A.R., 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74, 5463–5467. Sekine, Y., Ohtsubo, E., 1989. Frameshifting is required for production of the transposase encoded by insertion sequence 1. Proc. Natl. Acad. Sci. USA 86, 4609–4613. Sekine, Y., Ohtsubo, E., 1991. Translational frameshifting in IS elements and other genetic systems. In: Kimura, M., Takahata, N. ( Eds.), New Aspects of The Genetics of Molecular Evolution. Japan Scientific Societies Press, Tokyo/Springer-Verlag, Berlin, pp. 243–261. Sekine, Y., Nagasawa, H., Ohtsubo, E., 1992. Identification of the site of translational frameshifting required for production of the transposase encoded by insertion sequence IS1. Mol. Gen. Genet. 235, 317–324. Sekine, Y., Eisaki, N., Ohtsubo, E., 1994. Translational control in production of transposase and in transposition of insertion sequence IS3. J. Mol. Biol. 235, 1406–1420. Sekine, Y., Eisaki, N., Ohtsubo, E., 1996. Identification and characterization of the linear IS3 molecules generated by staggered breaks. J. Biol. Chem. 271, 197–202. Sekino, N., Sekine, Y., Ohtsubo, E., 1995. IS1-encoded proteins, InsA and the InsA-B∞-InsB transframe protein (transposase): functions deduced from their DNA-binding ability. Adv. Biophys. 31, 209–222. Silhavy, T.J., 1984. In: Experiments with Gene Fusions. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 11–12. Takeshita, S., Sato, M., Toba, M., Masahashi, W., Hashimoto-Gotoh, T., 1987. High-copy-number and low-copy-number plasmid vectors for lacZ a-complementation and chloramphenicol- or kanamycinresistance selection. Gene 61, 63–74. Turlan, C., Chandler, M., 1995. IS1-mediated intramolecular rearrangements: formation of excised transposon circles and replicative deletions. EMBO J. 14, 5410–5421. Umeda, M., Ohtsubo, E., 1991. Four types of IS1 with dierence in nucleotide sequences reside in the Escherichia coli K-12 chromosome. Gene 98, 1–5. Vieira, J., Messing, J., 1987. Production of single stranded plasmid DNA. Methods Enzymol. 153, 3–11. Wada, M., Kano, Y., Ogawa, T., Okazaki, T., Imamoto, F., 1988. Construction and characterization of the deletion mutant of hupA and hupB genes in Escherichia coli. J. Mol. Biol. 204, 581–591. Yoshioka, Y., Ohtsubo, H., Ohtsubo, E., 1987. Repressor gene finO in plasmids R100 and F: Constitutive transfer of plasmid F is caused by insertion of IS3 into F finO. J. Bacteriol. 169, 619–623.
Isolation and characterization of IS1 circles Yasuhiko Sekine, Naoki Eisaki, Kiyoaki Kobayashi, Eiichi Ohtsubo * Institute of Molecular and Cellular Biosciences, The University of Tokyo, Bunkyo-ku, Yayoi 1-1-1, Tokyo 113, Japan Received 16 August 1996; received in revised form 2 December 1996; accepted 14 December 1996; Received by A. Nakazawa
Abstract Transposase encoded by insertion sequence IS1 is produced from two out-of-phase reading frames by translational frameshifting that occurs in a run of adenines. An IS1 mutant with a single adenine insertion in the run of adenines eciently produces transposase, resulting in generation of miniplasmids by deletion for a region adjacent to IS1 from a plasmid carrying the IS1 mutant. Here, we found that besides miniplasmids, cells harboring the plasmid contained minicircles without the region required for replication. Cloning and DNA sequencing of the minicircles revealed that most of them were IS1 circles consisting of the entire IS1 sequence and a sequence, 5–9 bp in length, which intervenes between terminal inverted repeats, IRL and IRR, of IS1. Analysis of more IS1 circles isolated by polymerase chain reaction revealed that the intervening sequence was derived from the region flanking either IRL or IRR in the parental plasmid, suggesting that IS1 circles are generated by an excision event from the parental plasmid. The IS1 circles may be formed due to the cointegration reaction occurring within the parental plasmid carrying IS1. © 1997 Elsevier Science B.V. Keywords: Translational frameshifting; IS1 transposase; Minicircles; Cointegration; Simple insertion
1. Introduction Insertion sequence IS1 is the smallest active transposable element (768 bp) in bacteria and carries imperfect terminal inverted repeats, IRL and IRR, of about 30 bp long (Ohtsubo and Ohtsubo, 1978; Johnsrud, 1979). IS1, which appears as a simple insertion in bacterial and phage genes (Hirsch et al., 1972; Fiandt et al., 1972; Calos et al., 1978; Grindley, 1978 ), can mediate cointegration between a donor replicon carrying IS1 and a recipient replicon (Ohtsubo et al., 1980, 1981; Iida and Arber, 1980; Galas and Chandler, 1982). Simple insertion of IS1 is assumed to occur due to cleavage of the strands of the donor replicon in the strand-transfer product, an intermediate molecule of cointegration, followed by release and degradation of the donor portion (Ohtsubo et al., 1981). The cointegration reaction that occurs intramolecularly is believed to generate a deletion of a DNA segment adjacent to IS1 (Ohtsubo et al., * Corresponding author. Tel./fax: +81 3 56843269; e-mail: [email protected] Abbreviations: bp, base pair (s); IR(s), terminal inverted repeat (s); IRL and IRR, terminal inverted repeats at left and right, respectively; IS, insertion sequence; nt, nucleotide(s); R , resistant/resistance; Km, kanamycin; Cm, chloramphenicol; PCR, polymerase chain reaction. 0378-1119/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S0 3 78 - 11 19 ( 9 7 ) 00 05 7 -7
1978; Sekine and Ohtsubo, 1991; Sekino et al., 1995; Turlan and Chandler, 1995). Upon cointegration, IS1 gives rise to duplication of the target sequence, mostly 9 bp in length (Ohtsubo et al., 1980; Machida and Machida, 1987 ). IS1 encodes two open reading frames, insA and insB (Machida et al., 1982, 1984; Jakowec et al., 1988). A frameshifting event in the −1 direction from the 3∞-end region of insA to an open reading frame B∞-insB, in which B∞ is the frame extended from the 5∞ end of insB and overlaps insA in −1 phase, is involved in production of the InsA-B∞-InsB transframe protein that is IS1 transposase (Sekine and Ohtsubo, 1989; Escoubas et al., 1991). The frameshifting occurs in a run of six adenines in the overlapping region between insA and B∞-insB (Sekine et al., 1992). An IS1 mutant (IS1-31) with a single adenine insertion in the run of adenines, which results in in-frame alignment of insA and B∞-insB, eciently produces transposase without frameshifting. This mutant in a plasmid mediates cointegration at a frequency much higher than wild-type IS1 (Sekine and Ohtsubo, 1989). The mutant also mediates deletion to give rise to a large amount of miniplasmids from the plasmid carrying it (Sekine and Ohtsubo, 1991; Sekino et al., 1995 ). Here, we report that the plasmid carrying IS1-31 with
184
Y. Sekine et al. / Gene 191 (1997) 183–190
a single adenine insertion generates miniplasmids as well as characteristic IS1 circles, consisting of the entire IS1 sequence and a short intervening sequence between IRL and IRR of IS1. The IS1 circles identified here may correspond to an excised circular copy of the transposon with two end regions of IS1, which was observed in the cells where transposase was produced eciently ( Turlan and Chandler, 1995). We present models for generation of the IS1 circles.
2. Materials and methods 2.1. Bacterial strains and plasmids Bacterial strains used were Escherichia coli K12 derivatives, YK1100 ( Wada et al., 1988 ), MV1184 (Vieira and Messing, 1987 ), MC4100 (Silhavy, 1984 ), and NM554 (Raleigh et al., 1988 ). Plasmid pSEK131 used was a pUC119-derivative carrying IS1-31, an IS1 mutant with a single adenine insertion in the run of adenines (Sekine et al., 1992). Plasmid pSEK117 is a pUC119-derivative carrying wild type IS1 (Sekine et al., 1992). Plasmid pSEK131TI, which is a type I miniplasmid derived from pSEK131, is deleted for the region (973 bp) extending from the end of IRL to a site located upstream of the bla gene (Sekine et al., 1992). Plasmid pHSG398 ( Takeshita et al., 1987) is a pUC18-type plasmid used for cloning of minicircles. Plasmid DNA was prepared from bacterial cells grown in L-rich broth. Small-scale preparation of plasmid DNA was performed as described by He et al. (1990). The alkaline lysis method (Sambrook et al., 1989) was used to prepare plasmid DNA for cloning and nucleotide sequencing. 2.2. Media Culture media used were L broth, L-rich broth and w-medium ( Yoshioka et al., 1987). w-Medium was used for transformation of plasmid DNA. L-agar plates contained 1.5% (w/v) agar (Eiken) in L broth. Antibiotics were added in L-agar plates, when necessary, at the concentration of 100 mg ampicillin ( Wako)/ml, 30 mg kanamycin (Sigma)/ml, or 30 mg chloramphenicol (Sigma)/ml. 2.3. Nucleotide sequencing Nucleotide sequences were determined by the dideoxynucleotide method (Sanger et al., 1977; Messing, 1982) using the Sequenase DNA sequencing kit ( U.S. Biochemical Corp.). Oligonucleotides corresponding to the sequence at nt 70 to 51 of the lower strand (named IS1-L) and to the sequence at nt 704 to 723 of the upper strand (named IS1-R) of IS1 with coordinates 1–768
(Ohtsubo and Ohtsubo, 1978) were used as primer for sequencing. These oligonucleotides were synthesized using a DNA synthesizer model 392 (Applied Biosystems). The DNA chains were labeled with [a-32P]dCTP (15 TBq/mmol, Amersham) and separated on 8% polyacrylamide gels containing 7 M urea. 2.4. Cloning of IS1 circles IS1 circles were cloned as follows: Total closed circular DNA was prepared from each of two clones of the YK1100 cells harboring pSEK131 or each of two clones of the MV1184 cells harboring pSEK131. The DNA was digested with PstI (Takara), which cleaves pSEK131 at one site within IS1, and subjected to electrophoresis in a 5% polyacrylamide gel. The DNA (approximately 0.8 kb) eluted from the gel was inserted into the PstI site of pHSG398. The DNA sample was introduced into NM554 by transformation and the CmR transformants were selected. Plasmids in several CmR transformants originated from the DNA sample from a cell culture were prepared and analyzed for their structures by sequencing. Critical regions in the IS1 circles containing an intervening sequence were cloned as follows: The total closed circular DNA prepared from each of three clones of the MV1184/pSEK131 cells, each of four clones of the MC4100/pSEK131TI cells, or each of four clones of the NM554/pSEK131TI cells was digested with PstI. Using the DNA preparation (0.5 ng) as template and a pair of primers IS1-L and IS1-R (50 pmol each), polymerase chain reaction ( PCR) was carried out in 50 ml of Cetus buer containing 2.5 units of AmpliTaq DNA polymerase ( Perkin Elmer Cetus). Incubations at 93°C (2 min), 50°C (2 min), and 72°C (3 min) were repeated 28 times in an automated thermal cycler. The DNA fragment amplified was inserted into the SmaI site of pUC119, and the plasmid DNA sample was introduced into NM554 by transformation.
3. Results and discussion 3.1. Identification of IS1 circles in cells harboring pSEK131 Plasmid pSEK131 (3.9 kb in length) is a derivative of pUC119 carrying IS1-31. Total plasmid DNA molecules, which were prepared from cells of an E. coli strain YK1100(recA+) harboring this plasmid, contained miniplasmids of dierent sizes, approximately 2.8 and 2.0 kb in length (Fig. 1A), while those from cells harboring pSEK117, another pUC119-derivative carrying wild type IS1, did not (Fig. 1B). These miniplasmids are supposed to correspond to those, called type I and II, previously identified and characterized (see shadowed
Y. Sekine et al. / Gene 191 (1997) 183–190
185
Fig. 1. Ethidium bromide-stained gels showing production of miniplasmids and minicircles. (A) An agarose gel (0.7%) showing total closed circular DNA prepared from the YK1100 cells harboring pSEK117 or pSEK131. ( B) An agarose gel ( 0.7%) showing the two DNA preparations digested with PstI. (C ) A polyacrylamide gel ( 5%) showing the PstI digests of total closed circular DNA prepared from the YK1100 cells harboring pSEK131. Positions of the parental plasmid DNA, miniplasmids (type I and II ) and minicircles ( type III ) are indicated. Note that minicircles are not visualized in the agarose gels because of small amounts loaded.
boxes in Fig. 2), with a deletion extending from the IRL end of IS1 (Sekino et al., 1995 ). We then prepared a large amount of total plasmid DNA from the YK1100 cells harboring pSEK131. Electrophoresis in a polyacrylamide gel revealed that the DNA sample contained not only miniplasmids but also minicircular DNA molecules (here called type III; Fig. 2 ), approximately 0.8 kb in length, which could be linearized by digestion with PstI ( Fig. 1C). The minicircular DNA molecules digested with PstI were eluted from the gel, ligated with the PstI fragment containing the kanamycin-resistance ( KmR) gene, and introduced into YK1100 by transformation. We obtained no KmR transformants, showing that the minicircles, unlike miniplasmids (Sekino et al., 1995 ), cannot replicate autonomously due to lack of the region required for DNA replication. To analyze structures of the minicircles, the minicircular DNA molecules digested with PstI were inserted into the PstI site of a chloramphenicol-resistance (CmR) plasmid pHSG398. The minicircular DNA molecules were also prepared from cells of another E. coli strain MV1184 (recA−) harboring pSEK131 and inserted into pHSG398. DNA sequencing analysis of plasmids in the CmR transformants revealed that eleven out of seventeen clones obtained contained minicircles with a structure consisting of the entire IS1-31 sequence and an intervening sequence (6–9 bp in length) between IRL and IRR of IS1 ( Figs. 2 and 3A). We call these minicircles as IS1 circles, since they were distinct from the other minicircles (see legend to Fig. 2 ). Of eleven clones of the IS1 circles,
five clones had an intervening sequence identical to that flanking either IRR or IRL of IS1 in the parental plasmid pSEK131 (see clones 7, 9, M6, M8, and M10 in Fig. 3A). Four clones had an intervening sequence which was present at a region distant from IS1 in pSEK131 (see clones 4, 6, M1, and M9 in Fig. 3A). The intervening sequences in the other two clones were, however, not found in pSEK131 (see clones 13 and 31 in Fig. 3A). There was no significant structural dierence among the IS1 circles generated in both host strains (recA+ and recA−) used here, suggesting that the RecA function is not involved in formation of the IS1 circles.
3.2. Characterization of IS1 circles To characterize the IS1 circles, we examined more IS1 circles by isolating their portions by polymerase chain reaction (PCR) using total plasmid DNA prepared from independent cultures of the MV1184 cells harboring pSEK131 as template and a pair of primers that hybridize with IRL and IRR and prime DNA synthesis toward outside of IS1. DNA sequencing analysis of the amplified fragments cloned into the SmaI site of pUC119 revealed that all the clones from each of three samples from independent cultures contained the fragments with a structure consisting of two end regions of IS1 and an intervening sequence between IRL and IRR, 6–9 bp in length (Fig. 3B). Five clones contained the intervening sequence identical to that flanking IRR of IS1, whereas
186
Y. Sekine et al. / Gene 191 (1997) 183–190
Fig. 2. Structures of miniplasmids and minicircles derived from plasmid pSEK131. pSEK131 is a pUC119-derivative carrying IS1-31 with IRL and IRR (triangles) and is shown in a linear representation. Italic letters are coordinates given to pSEK131 (1–3926) with IS1-31 ( 898–1666). The coding region for the bla gene (a thick arrow) and a primer for DNA replication, RNA II (thin arrow), are shown. ori, the replication origin of pUC119. A restriction site for PstI located within IS1 is shown. The nucleotide sequence of a region in IS1-31 containing the run of adenines with a single adenine insertion, which results in in-frame fusion between insA and B∞-insB, is shown. Shaded thick lines represent approximate structures of miniplasmids (type I and II ) and minicircles (type III ). The type I and II miniplasmids have a sequence starting from the exact end at IRL of IS1-31 to a site within a restricted region indicated at the right side of the structures of miniplasmids (Sekino et al., 1995). Thin lines indicate structures of the cloned minicircles. Numbers with and without M indicate minicircles derived from pSEK131 in MV1184 and YK1100, respectively. IS1 circles among minicircles had a unique structure with an intervening sequence which is shown in Fig. 3A. Of the other minicircles, two are those with the entire IS1 sequence and a pUC119 sequence, 35 or 59 bp in length, which is much longer than those in the IS1 circles (see clones 10 and 12). Four are those with IS1 which lacked a terminal region of either IRL or IRR at positions indicated (see clones 1, M2, M7, and M3). These minicircles are supposed to be the products of the deletion extending from one end of IS1-31 to a site near or within the other IS1-31 end region.
another five contained the intervening sequence identical to that flanking IRL of IS1, and one clone contained a 9-bp sequence identical to that duplicated at both ends of IS1 (Fig. 3B). The intervening sequence in another IS1 circle, however, was found at a region distant from IS1 in the parental plasmid pSEK131 (Fig. 3B). Among the IS1 circles identified in this or the previous section, some had an intervening sequence which is identical to the sequence flanking IRL or IRR of IS1 in the parental plasmid. This suggests that they are formed
directly from the parental plasmid by an excision event. The other IS1 circles, however, had an intervening sequence (shown by thin letters in Fig. 3 ) which is dierent from the sequence flanking IRL or IRR and from one another. Many of such intervening sequences were found in the sequence of the parental plasmid at the regions near the points of deletion to give rise to type I and II miniplasmids (see Figs. 2 and 3 ). This suggests that these IS1 circles are generated from miniplasmids present in cells harboring pSEK131.
Y. Sekine et al. / Gene 191 (1997) 183–190
187
Fig. 4. Nucleotide sequences intervening between IRL and IRR in the IS1 circles derived from a miniplasmid pSEK131TI in MC4100 or NM554. A critical portion of pSEK131TI is shown at top. Note that the 9-bp sequences ( boldface letters) flanking IS1-31 are dierent each other. For the other information, see legend to Fig. 3.
3.3. IS1 circles generated from a miniplasmid
Fig. 3. Structures of the IS1 circles with a sequence intervening between IRL and IRR. A critical portion of the parental plasmid pSEK131 containing IS1-31 with IRL and IRR ( triangles) and the flanking regions is shown at top. A 9-bp sequence ( boldface letters) is duplicated at the regions adjacent to IS1-31. The IS1 circles contain a nucleotide sequence, 6–9 bp in length, that intervenes between IRL and IRR. The IS1 circles are shown in a linear representation by cutting them at the point between an IR and the intervening sequence. The IS1 circles shown in ( A) are depicted schematically in Fig. 2. The IS1 circles shown in (B) were actually obtained as the fragments containing the intervening sequences, which were amplified by PCR and inserted into pUC119 (see section 2: Materials and methods). These IS1 circles have numbers, 1–3, which indicate clones isolated from independent cultures of the MV1184/pSEK131 cells. Alphabetic letters (a–e) are clones from the same culture of cells. The intervening sequences, which are identical to a sequence flanking an IR of IS1-31, are indicated by boldface uppercase letters. The intervening sequences indicated by the non-boldface uppercase letters are present in pSEK131 at positions as indicated in parentheses (see map of pSEK131 in Fig. 2). The intervening sequences in clones 4, M1, M9 and 1a are positioned on the side of IRL, because the IS1 circles with these sequences are supposed to be derived from miniplasmids with a deletion extending from IRL, whereas the intervening sequence in clone 6 is positioned on the side of IRR, because the IS1 circle with the sequence is supposed to be
Plasmid pSEK131TI is a type I miniplasmid derived from pSEK131 by deletion for the region ( 973 bp) extending from the end of IRL to a site located upstream of the bla gene (see Fig. 2 for the structure of pSEK131). pSEK131TI thus has an IRL-flanking sequence dierent from that in the parental plasmid (see Fig. 4), but still generates smaller miniplasmids and the IS1 circles (data not shown). To analyze the IS1 circles, we carried out PCR using total plasmid DNA prepared from either MC4100 (recA+) or NM554 (recA−) harboring pSEK131TI as template and a pair of primers that hybridize with IRL and IRR, and obtained 21 clones of the fragments containing the intervening sequences in the IS1 circles. DNA sequencing revealed that the fragments in the clones consisted of two end regions of IS1 and an intervening sequence between IRL and IRR with 5–8 bp in length (Fig. 4 ). Among them, ten clones contained the intervening sequence identical to that derived from a miniplasmid with a deletion extending from IRR. The intervening sequences shown by lowercase letters in clones 13 and 31 are not present in pSEK131. Note that positioning of the intervening sequences on the side of IRR is tentative, because the IS1 circles with these sequences are supposed to be derived from the E. coli chromosome where IS1-31 had been inserted.
188
Y. Sekine et al. / Gene 191 (1997) 183–190
flanking IRR in the parental plasmid pSEK131TI, whereas one contained the intervening sequence identical to that flanking IRL ( Fig. 4 ), supporting the previous suggestion that these are formed by an excision event from the miniplasmids. The intervening sequences in seven clones were not identical to those flanking IS1-31, but were found in pSEK131TI near the junction points of deletion almost exclusively in the type II miniplasmids ( Fig. 4 ). This suggests that these IS1 circles are generated by excision from miniplasmids present in cells harboring pSEK131TI. Three clones contained the intervening sequences, which were not found in pSEK131TI (Fig. 4 ). Such clones have been also found in the IS1 circles previously identified as described above. We assume that these are chromosomal sequences which became adjacent to IS1-31 transposed from pSEK131TI (or miniplasmids), and then were incorporated into the IS1 circles. It should be noted here that these IS1 circles were not derived from the IS1 elements already resident in the E. coli chromosome ( Umeda and Ohtsubo, 1991), since the nucleotide sequences of the intervening sequences in the IS1 circles were not the same as those flanking each IS1 element in the chromosome, and since the internal regions of several IS1 circles examined were found to contain an adenine insertion characteristic of IS1-31. Here again, there was no significant structural dierence between the IS1 circles generated in both strains (recA+ and recA−), confirming that the RecA function is not involved in formation of the IS1 circles. 3.4. Possible models for formation of the IS1 circles How are the IS1 circles formed? IS1 forms cointegrates between a donor molecule carrying IS1 and a recipient molecule. In the first step of cointegration, each strand of the donor molecule is assumed to be cut at the 3∞ end of IS1 to generate a pair of 3∞-OH termini, each of which is transferred to each strand at a target site sequence in the recipient molecule. The IS1 circles may be formed due to the cointegration reaction occurring in the donor molecules, such that one of the 3∞-OH termini is transferred to the same strand of a donor molecule at a site in the sequence flanking another end of IS1, followed by DNA synthesis initiated from the other 3∞-OH terminus to fill the gap in the opposite strand and by cleavage to remove the non-IS1 sequence (model A in Fig. 5 ). Alternatively, each of the 3∞-OH termini might be transferred to a site in the sequence flanking the other end of IS1. If the sequences flanking IS1 are dierent from each other, the intervening sequence in the IS1 circle will be a heteroduplex form (model B in Fig. 5 ). It is also possible that in the initial step of cointegration, only one strand happens to be cut at the 3∞ end of IS1, and that the 3∞-OH terminus is transferred to the sequence flanking the other IS1 end,
leading to formation of an IS1 circle via cleavage of a strand to give the 3∞ end used for repair synthesis (model C in Fig. 5). Although we have no experimental evidence to determine which model is the case, these reactions may occur in the DNA-protein complex in which both IRs are held together by transposase; in such a situation, the 3∞-OH terminus at one end generated by transposase is located near the sequence flanking another end of IS1, which is the target of its attack, to result in generation of the IS1 circles with a short intervening sequence, 5–9 bp in length. We have observed in this study that the IS1 circles in some clones isolated independently had the same intervening sequence (Figs. 3 and 4). This suggests that some sites are more preferentially used as the target for the attack of the 3∞-OH terminus to generate the IS1 circles than other sites. Turlan and Chandler (1995) have recently reported that excised circular molecules of the transposon with two end regions of IS1 are present in the cells where transposase is produced eciently. These molecules had an intervening sequence like the IS1 circles identified here. The excised circular molecules, however, had the same intervening sequence, 8 bp in length, in the region flanking IRR of IS1 (Turlan and Chandler, 1995). This suggests that only one particular site has been used as the target for the attack of the 3∞-OH terminus to result in generation of the excised transposon molecules. IS3 and IS911 are closely related elements, which are distinct from IS1, since they transpose in a nonreplicative manner and thus do not form cointegrates. These elements, however, can generate characteristic IS circles consisting of the entire IS sequence and a 3-bp sequence intervening between both IRs, when transposases are eciently produced ( Polard et al., 1992; Sekine et al., 1994). In the case of IS911, strand-transfer products depicted in model C in Fig. 5 are observed, leading to the assumption that such molecules are intermediates of the IS911 circles (Polard and Chandler, 1995). 3.5. Role of the IS circles What is the role of IS circles? We have observed that IS3 generates linear molecules with a pair of 3∞-OH termini of IS3 (Sekine et al., 1996 ). We have also observed that the IS3 circles are linearized by a staggered break introduced in the junctions between the intervening sequence and either IRL or IRR of IS3 (Aihara, K., Sekine, Y. and Ohtsubo, E., unpublished results). It has been proposed that the linear molecules are subsequently transferred to a target molecule to generate a simple insertion product like integration of retroviruses upon retroposition (Sekine et al., 1996 ). In IS1, linear molecules have not, however, found, suggesting that IS1 circles can transpose in a dierent mode from IS3 not through linearization of the circles, or that the IS1
Y. Sekine et al. / Gene 191 (1997) 183–190
189
Fig. 5. Possible models ( A–C) for formation of the IS1 circles. Two strands of IS1 are shown by thick lines. Small arrows indicate positions of a nick introduced in a DNA strand. Broken arrows indicate transfer of the 3∞-OH termini. Wavy arrows indicate displacement DNA synthesis.
circles are the end products, which are inert for transposition. Further experiments are required to elucidate the role of the IS1 circles.
Acknowledgement This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan.
4. Conclusions (1 ) E. coli cells harboring a plasmid with an IS1 mutant, which eciently produces transposase, contained minicircles lacking the region required for replication. Most of them were IS1 circles with a characteristic structure with the entire IS1 sequence and a sequence, 5–9 bp long, intervening between IRL and IRR of IS1. (2 ) Most of the intervening sequences in the IS1 circles are identical to the sequence flanking either IRL or IRR in the parental plasmid, indicating that such IS1 circles were generated by an excision event from the parental plasmid. (3 ) The IS1 circles may be formed due to the cointegration reaction occurring in a parental plasmid containing IS1.
References Calos, M.P., Johnsrud, L., Miller, J.H., 1978. DNA sequence at the integration sites of the insertion element IS1. Cell 13, 411–418. Escoubas, J.M., Pre`re, M.F., Fayet, O., Salvignol, I., Galas, D., Zerbib, D., Chandler, M., 1991. Translational control of transposition activity of the bacterial insertion sequence IS1. EMBO J. 10, 705–712. Fiandt, M., Szybalski, W., Malamy, M.H., 1972. Polar mutations in lac, gal, phage l consist of a few IS-DNA sequences inserted with either orientation. Mol. Gen. Genet. 119, 223–231. Galas, D.J., Chandler, M., 1982. Structure and stability of Tn9-mediated cointegrates. Evidence for two pathways of transposition. J. Mol. Biol. 154, 245–272. Grindley, N.D.F., 1978. IS1 insertion generates duplication of a nine base pair sequence at its target site. Cell 13, 419–426. He, M., Wilde, A., Kaderbhai, M.A., 1990. A simple single-step
190
Y. Sekine et al. / Gene 191 (1997) 183–190
procedure for small-scale preparation of Escherichia coli plasmid. Nucleic Acids Res. 18, 1660. Hirsch, H., Starlinger, P., Brachet, P., 1972. Two kinds of insertions in bacterial genes. Mol. Gen. Genet. 119, 191–206. Iida, S., Arber, W., 1980. On the role of IS1 in the formation of hybrids between the bacteriophage P1 and the R plasmid NR1. Mol. Gen. Genet. 177, 261–270. Jakowec, M., Prentki, P., Chandler, M., Galas, D., 1988. Mutational analysis of the open reading frames in the transposable element IS1. Genetics 120, 47–55. Johnsrud, L., 1979. DNA sequence of the transposable element IS1. Mol. Gen. Genet. 169, 213–218u Machida, Y., Machida, C., Ohtsubo, E., 1984. Insertion element IS1 encodes two structural genes required for its transposition. J. Mol. Biol. 177, 229–245. Machida, C., Machida, Y., 1987. Base substitutions in transposable element IS1 cause DNA duplication of various length at the target site for plasmid cointegration. EMBO J. 6, 1799–1803. Machida, Y., Machida, C., Ohtsubo, H., Ohtsubo, E., 1982. Factors determining frequency of plasmid cointegration mediated by insertion sequence IS1. Proc. Natl. Acad. Sci. USA 79, 277–281. Messing, J., 1982. New M13 vectors for cloning. Methods Enzymol. 101, 20–78. Ohtsubo, H., Ohtsubo, E., 1978. Nucleotide sequence of an insertion element, IS1. Proc. Natl. Acad. Sci. USA 75, 615–619. Ohtsubo, E., Rosenbloom, H., Schrempf, H., Goebel, W., Rosen, J., 1978. Site specific recombination involved in the generation of small plasmids. Mol. Gen. Genet. 159, 131–141. Ohtsubo, E., Zenilman, M., Ohtsubo, H., 1980. Plasmids containing insertion elements are potential transposons. Proc. Natl. Acad. Sci. USA 77, 750–754. Ohtsubo, E., Zenilman, M., Ohtsubo, H., McCormick, M., Machida, C., Machida, Y., 1981. Mechanism of insertion and cointegration mediated by IS1 and Tn3. Cold Spring Harbor Symp. Quant. Biol. 45, 283–295. Polard, P., Chandler, M., 1995. An in vivo transposase-catalyzed single-stranded DNA circularization reaction. Genes Dev. 9, 2846–2858. Polard, P., Pre`re, M.F., Fayet, O., Chandler, M., 1992. Transposaseinduced excision and circularization of the bacterial insertion sequence IS911. EMBO J. 11, 5079–5090. Raleigh, E.A., Murray, N.E., Revel, H., Blumenthal, R.M., Westaway, D., Reith, A.D., Rigby, P.W., Elhai, J., Hanahan, D., 1988. McrA and McrB restriction phenotype of some E. coli strains and implications for gene cloning. Nucleic Acids Res. 16, 1563–1574. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 1.25–1.28.
Sanger, F., Nicklen, S., Coulson, A.R., 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74, 5463–5467. Sekine, Y., Ohtsubo, E., 1989. Frameshifting is required for production of the transposase encoded by insertion sequence 1. Proc. Natl. Acad. Sci. USA 86, 4609–4613. Sekine, Y., Ohtsubo, E., 1991. Translational frameshifting in IS elements and other genetic systems. In: Kimura, M., Takahata, N. ( Eds.), New Aspects of The Genetics of Molecular Evolution. Japan Scientific Societies Press, Tokyo/Springer-Verlag, Berlin, pp. 243–261. Sekine, Y., Nagasawa, H., Ohtsubo, E., 1992. Identification of the site of translational frameshifting required for production of the transposase encoded by insertion sequence IS1. Mol. Gen. Genet. 235, 317–324. Sekine, Y., Eisaki, N., Ohtsubo, E., 1994. Translational control in production of transposase and in transposition of insertion sequence IS3. J. Mol. Biol. 235, 1406–1420. Sekine, Y., Eisaki, N., Ohtsubo, E., 1996. Identification and characterization of the linear IS3 molecules generated by staggered breaks. J. Biol. Chem. 271, 197–202. Sekino, N., Sekine, Y., Ohtsubo, E., 1995. IS1-encoded proteins, InsA and the InsA-B∞-InsB transframe protein (transposase): functions deduced from their DNA-binding ability. Adv. Biophys. 31, 209–222. Silhavy, T.J., 1984. In: Experiments with Gene Fusions. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 11–12. Takeshita, S., Sato, M., Toba, M., Masahashi, W., Hashimoto-Gotoh, T., 1987. High-copy-number and low-copy-number plasmid vectors for lacZ a-complementation and chloramphenicol- or kanamycinresistance selection. Gene 61, 63–74. Turlan, C., Chandler, M., 1995. IS1-mediated intramolecular rearrangements: formation of excised transposon circles and replicative deletions. EMBO J. 14, 5410–5421. Umeda, M., Ohtsubo, E., 1991. Four types of IS1 with dierence in nucleotide sequences reside in the Escherichia coli K-12 chromosome. Gene 98, 1–5. Vieira, J., Messing, J., 1987. Production of single stranded plasmid DNA. Methods Enzymol. 153, 3–11. Wada, M., Kano, Y., Ogawa, T., Okazaki, T., Imamoto, F., 1988. Construction and characterization of the deletion mutant of hupA and hupB genes in Escherichia coli. J. Mol. Biol. 204, 581–591. Yoshioka, Y., Ohtsubo, H., Ohtsubo, E., 1987. Repressor gene finO in plasmids R100 and F: Constitutive transfer of plasmid F is caused by insertion of IS3 into F finO. J. Bacteriol. 169, 619–623.