Development of an intermolecular transposition assay system in Bacillus subtilis 168 using IS4Bsu1 from Bacillus subtilis (natto)

Transposons are defined as DNA elements that can transpose from one genomic location to another. They generally have terminal inverted repeat (IR) sequences and a transposase gene that mediates transposition and generates the duplication of the target site sequence that accompanies transposition (Mahillon & Chandler, 1998). Insertion sequences (ISs) are small and simple transposons in bacteria; they are components of chromosomes and plasmids. Many IS elements have been identified in various bacteria and classified on the basis of transposase homology, IRs and the length of the target sequence. It is widely believed that transposition of IS elements plays an important role as a motive force in genomic evolution (Craig et al., 2002). The transposition activity of transposons is mediated by various host factors that are specific for each element. Histone-like proteins (or DNA chaperones) such as HU and integration host factor (IHF) function in the transposition reaction of some bacterial transposons (Kleckner et al., 1996; Lavoie & Chaconas, 1996). While a nucleoid-associated protein, H-NS, is required for IS1 transposition, other histone-like proteins are not (Shiga et al., 2001). Therefore, there is no particular host factor essential for the transposition of all transposons. Other known host factors include the replication initiator DnaA, the protein chaperones/proteases ClpX, ClpP and ClpA, the SOS control protein LexA and the Dam methylase (Mahillon & Chandler, 1998).

Natto is a traditional fermented food derived from soybeans; it is widely consumed by the Japanese. A string-like component contributes to its characteristic sticky texture; it has been found to be a fermentation product, poly-γ-glutamate (γ-PGA), of a variant strain of Bacillus subtilis (natto). B. subtilis (natto) is closely related to B. subtilis Marburg 168, the best-characterized Gram-positive bacterium, whose entire genome has been sequenced (Kunst et al., 1997). Among B. subtilis strains, Marburg 168 can be transformed at a high frequency with its own natural genetic competence; other B. subtilis strains, including B. subtilis (natto), yield transformants at a frequency as low as that of spontaneous mutation.

In the B. subtilis 168 strain, cell-density-dependent phenotypes are regulated by a quorum-sensing mechanism involving the ComP–ComA two-component regulatory system (Hahn & Dubnau, 1991; Nakano et al., 1991; Roggiani & Dubnau, 1993; Solomon et al., 1995; Weinrauch et al., 1990). The synthesis of γ-PGA in B. subtilis (natto) is also controlled by this system. The ability to produce γ-PGA is occasionally lost through serial cultivation, and this phenomenon is associated with the transposition of an IS into the comP gene (Nagai et al., 2000). The IS found in natto strains is a member of the IS4 family, designated IS4Bsu1, which is 1406 bp in length and has imperfect 18 bp terminal IRs. It also contains an ORF that encodes a 374 aa transposase and generates a 9 bp duplication of the target site during insertion (Nagai et al., 2000).

Transposition studies have focused on analysing the detailed mechanisms of transposition; however, the cellular conditions that induce transposition remain to be elucidated. Therefore, we constructed a transposition assay system using B. subtilis 168 and modified IS4Bsu1. Our results revealed an increase in transposition frequency under high-temperature and competence-developing conditions.

Bacterial strains, plasmids, and media.
Assay strains NBS040 (trpC2 amyE : : IRL-cat-IRR-P_S10-tnp), NBS041 (trpC2 amyE : : cat-IRR-P_S10-tnp), and NBS042 (trpC2 amyE : : IRL-cat-IRR), all B. subtilis Marburg 168 (trpC2) derivatives, were constructed as described below. A transformable natto strain RIK7102 [bio mecA : : spc amyE : : comG–lacZ (Cm^R)] (Ashikaga et al., 2000) was obtained from F. Kawamura (Rikkyo University). These strains were grown in Luria–Bertani (LB) or CI medium (Anagnostopoulos & Spizizen, 1961) at 37 or 49 °C. Escherichia coli DH10B [F^– mcrA Δ(mrr-hsdRMS-mcrBC) φ80dlacZ ΔM15 ΔlacX74 deoR recA1 araD139 Δ(ara leu)7697 galU galK λ^– rpsL endA1 nupG] was grown in LB medium at 37 °C. The shuttle vector pDG148 [ampicillin resistant (Amp^R) for E. coli and kanamycin resistant (Km^R) for B. subtilis] (Stragier et al., 1988) was used for the intermolecular transposition assay. When necessary, biotin was added at a final concentration of 0.1 µg ml^–1. Antibiotics were used at the following concentrations: chloramphenicol, 5 µg ml^–1 (B. subtilis) or 20 µg ml^–1 (E. coli); ampicillin, 50 µg ml^–1; kanamycin, 5 µg ml^–1; and spectinomycin, 50 µg ml^–1.

Southern blot analysis.
For genomic Southern blot analysis, 6 µg chromosomal DNA was digested with EcoRV (TaKaRa) and resolved on an 0.8 % agarose gel. The DNA was then transfer-blotted onto a Hybond-N⁺ nylon membrane (Amersham Bioscience). Hybridization was with a DIG-labelled (DIG High Prime DNA labelling and detection starter kit II; Roche Diagnostics) IS4Bsu1-specific probe using the PCR-generated fragment (primers ISin3' and ISin5', Table 1) according to the manufacturer's instructions.

Table 1. Primers used in this study

Construction of strains for intermolecular transposition assay.
A mini-IS fragment containing a chloramphenicol resistance (Cm^R) cassette (cat) was inserted between the 40 bp left inverted repeat and the right inverted repeat (IRL and IRR, respectively), with the transposase gene (tnp) placed under the control of the S10 operon promoter (P_S10) positioned separately within the amyE locus. A strain carrying this mini-IS fragment was constructed as follows. First, five primary PCR-generated fragments were amplified. The PCR products (fragments 1 and 5) contained an 800 bp fragment of the amyE N-terminal region (generated using primers P1 and P2; Table 1) and the C-terminal region (primers P9 and P10), respectively. Fragment 2 included both the IRL and IRR of IS4Bsu1 and a cat gene (primers P3 and P4). Fragment 3 carried P_S10 and the Shine–Dalgarno sequence of the rpsJ gene of B. subtilis (primers P5 and P6). Fragment 4 (primers P7 and P8) included the tnp gene derived from IS4Bsu1. Templates for generating these fragments were B. subtilis 168 genomic DNA (fragments 1, 3 and 5), pBEST4C (Itaya et al., 1990) (fragment 2), and B. subtilis (natto) OK2 (Ashikaga et al., 2000) genomic DNA (fragment 4). Primers were designed to have 5' add-on sequences to create overlapping sequences between flanking fragments. To generate the desired secondary PCR constructs, fragment 6 (primers P11 and P12) was amplified by the recombinant PCR method, using the three-piece primary PCR products (fragments 2, 3 and 4) (Higuchi, 1989). Likewise, the tertiary PCR product (primers P1 and P10), generated from fragments 1, 5 and 6, was used to transform B. subtilis 168 to construct the assay strain NBS040 (Fig. 2). As control strains, the IRL-deleted strain NBS041 (using primers P15 and P16 instead of P3 and P2) and the tnp-depleted strain NBS042 (using primers P13 and P14 instead of P4 and P9) were constructed using NBS040 genomic DNA as template (Fig. 2). Pyrobest DNA polymerase (TaKaRa) was used for all PCR assays.

(24K): Fig. 2. Schematic organization of the amyE locus of the B. subtilis 168 strains used for intermolecular transposition assays. Strain NBS040 carries a mini-IS and a PS10-driven tnp gene, whereas NBS041 lacks the IRL, and NBS042 lacks the tnp gene. The mini-IS carries the cat gene (white arrow), which is flanked by the IRL (right-directed black arrow head) and IRR (left-directed black arrow head) regions of IS4Bsu1. PS10, S10 promoter (thin arrow); tnp, transposase gene encoded by IS4Bsu1 (grey arrow); amyE-N and amyE-C, N and C-terminal regions of the amyE gene (light-grey box and arrow, respectively).

Detection of mini-IS transposition.
Plasmid pDG148, the target for transposition of the mini-IS, was introduced into NBS040, NBS041 and NBS042 cells, and transformants, selected on LB plates containing kanamycin, were then grown for 16 h under various conditions (see above). The initial OD₆₀₀ was adjusted to 0.01 (LB medium) or 0.003 (CI medium) to normalize the total generation number until full growth was reached for each culture condition. Plasmids were then extracted from the cultured cells by alkaline lysis (Harwood & Cutting, 1990), and 0.5 µg plasmid DNA was introduced into E. coli DH10B cells via electroporation with a Micro Pulser (Bio-Rad). After appropriate dilution, cells were plated on LB plates containing 50 µg ampicillin ml^–1 with or without 20 µg chloramphenicol ml^–1. The transposition frequency was estimated as the ratio of the number of Amp^R/Cm^R transformants to the number of total Amp^R transformants.

DNA sequencing.
The DNA sequence of the target duplication and the target sites of the mini-IS transposition into the pDG148 plasmid (Fig. 3, Table 4) were analysed using BigDye Terminator version 3.1 cycle sequencing kits and the GeneAmp PCR system 2700 (Applied Biosystems). The primers were out-cat3up and out-cat4down (Table 1); they anneal to the N- and C-terminal regions of the cat gene, respectively.

(28K): Fig. 3. Structure of the target plasmid pDG148 and insertion sites of mini-IS. The short arrows towards the plasmid indicate insertion sites and the numbers are clone numbers. Directions of mini-IS insertion are indicated: (+), orientation of the cat gene in the mini-IS is the same as that of the lacI gene on pDG148; (–), cat and lacI genes have opposite orientations. Note that many insertion sites were mapped downstream of the Pspac promoter and the nucleotide sequence of this region is shown enlarged.

Table 4. Target site sequences

High transposition frequency of IS4Bsu1 in competence-inducing media
During the course of studies to determine the loci of IS4Bsu1 in the natto genome, we observed that cells that underwent a transformation event appeared to have more copies of IS4Bsu1. To better understand this observation, we examined cells derived from natto strain RIK7102. Serial cultivation in LB or CI medium was used for transformation. After 80 generations, chromosomal DNA was extracted from each culture and the distribution of IS4Bsu1 was examined by Southern blot analysis (Fig. 1). While two extra copies of IS4Bsu1 were clearly identified in the genome of cells grown in CI medium, the pattern of the genome of cells grown in LB medium was identical to that of the initial strain, RIK7102, indicating that the transposition frequency of IS4Bsu1 was exceptionally high in competence-inducing medium. We also determined some of the loci of IS4Bsu1 in the genome of RIK7102; the loci that correspond to those in strain 168 are summarized in Table 2. Based on these observations, we posit that the localization of these loci is random and that transposition events are related to culture conditions.

(59K): Fig. 1. Southern blot analysis of IS4Bsu1 insertions in B. subtilis (natto) RIK7102. B. subtilis (natto) RIK7102 cells were grown at 37 °C to OD600 0.7 in LB medium. The culture was then diluted in LB or CI medium and cultivated for 80 generations. Chromosomal DNA from wild-type B. subtilis (natto) RIK7102 before serial cultivation (lane 1), from RIK7102 cultivated in competence medium for 80 generations (lane 2) and from RIK7102 cultivated in LB medium for 80 generations (lane 3) was digested with EcoRV and analysed by Southern blotting, as described in Methods. The two extra bands of IS4Bsu1 are indicated by arrowheads to the right of lane 2.

Table 2. Distribution of IS4Bsu1 in the B. subtilis (natto) genome

High transposition frequency of mini-IS under competence-developing and high-temperature conditions
As our observations suggested certain environmental conditions under which the transposition of IS4Bsu1 is induced, we postulated that molecular analysis of transposition induction using the genome-sequenced strain B. subtilis 168, instead of the natto strain, would shed light on this issue. Accordingly, we constructed an intermolecular transposition assay system based on B. subtilis 168 to monitor the behaviour of IS4Bsu1 under different environmental conditions (Fig. 2). Using this assay system and the detection techniques described in Methods, we were able to monitor quantitatively the transposition frequencies of mini-IS from the chromosome in the target plasmid pDG148.

To assess whether the transposition frequency of mini-IS changed under different conditions, we performed an intermolecular transposition assay. First we examined the mini-IS transposition frequency under normal culture conditions (LB medium at 37 °C). We found that the transposition frequency increased 15.7-fold when cells were grown under conditions leading to genetic competence induction, e.g. in CI medium, as compared to normal culture conditions (Table 3). In addition, cultivation in LB medium at 49 °C rather than 37 °C yielded a 4.4-fold higher transposition frequency. With strains NBS041 and NBS042 we obtained no Amp^R Cm^R transformants under any of the tested conditions. Our results indicate that the appearance of Cm^R transformants depends on the transposase and inverted repeat of IS4Bsu1 and that the phenomena we observed were a consequence of mini-IS transposition events.

Table 3. Mini-IS transposition frequency

Transposition of mini-IScat and IS4Bsu1 occurs in the same manner
DNA sequence analyses revealed that plasmids in Amp^R/Cm^R transformants carried insertions of mini-IS in either orientation at various sites in pDG148 (Fig. 3). Although the identification of 9 bp target sequence duplications at the insertion sites was evidence for the occurrence of IS transposition, the 9 bp random target sequences at different sites, usually AT-rich, were not conserved (Table 4). These results further demonstrated that our assay system reflected the ability of IS4Bsu1 to transpose into AT-rich random sequences and to yield 9 bp target sequence duplications. We developed an intermolecular transposition assay system using B. subtilis 168 and modified IS4Bsu1 to facilitate measurement of the transposition frequency (Fig. 2). We identified 9 bp target sequence duplications on both sides of the mini-IS (Table 4), indicating that our assay system actually detected phenomena resulting from transpositional mechanism(s). We first established that the transposition frequency of IS4Bsu1 in B. subtilis (natto) was remarkably high under competence-developing conditions (Fig. 1), suggesting the existence of some conditions that promote IS4Bsu1 transposition, including conditions that can lead to activation of host factors for transposition. Such environmental conditions have been described elsewhere (Eichenbaum & Livneh, 1998; Nagy & Chandler, 2004; Ohtsubo et al., 2005). Although Nagai et al. (2000) identified six to 11 copies of IS4Bsu1 in the B. subtilis (natto) genome, no transposons were found in the genome of the closely related B. subtilis 168 strain (Kunst et al., 1997; Qiu et al., 2003, 2004). Accordingly, we postulated that an intermolecular transposition assay system using B. subtilis 168 would facilitate simple and detailed transposition analysis.

Our intermolecular transposition assay revealed that under typical culture conditions (37 °C, LB medium; 1.4x10^–8 per target plasmid) the frequency of mini-IS transposition was low. Bacterial transposition activity is generally maintained at a low level because high transposition activity and the accompanying mutagenic effects of genomic rearrangement would be detrimental to the host cell (Mahillon & Chandler, 1998). We noted that the transposition frequency of mini-IS increased dramatically (15.7-fold) under our competence-developing (37 °C, CI medium) conditions and moderately (4.4-fold) under the high-temperature (49 °C, LB medium) conditions (Table 3). With respect to transposition frequency at high temperature (42 °C), contrasting results have been reported; it is higher in Burkholderia multivorans ATCC 17616 cells (Ohtsubo et al., 2005) and lower in E. coli cells (Nagy & Chandler, 2004).

Our results provide what is believed to be the first evidence for a high transposition frequency of mini-IS under competence-developing conditions. In B. subtilis, many different regulation pathways constitute the gene regulatory network that controls the development of competence (Hamoen et al., 2003). In particular, the competence transcription factor ComK activates expression of various genes involved in DNA binding, uptake and recombination (van Sinderen et al., 1995). The apparent high transposition frequency under competence-developing conditions in our study raises the possibility that in B. subtilis there are links between the regulatory pathways involved in competence development and transpositional events.

It is of note that these transposition phenomena could be achieved by expressing a P_S10-regulated tnp gene on the chromosome. The overexpression of Tn5 transposase in E. coli reportedly results in filamentation, aberrant nucleoid segregation and cell death (Weinreich et al., 1994). Similarly, in B. subtilis we noted filamentation and growth inhibition when we tried to construct a plasmid carrying the P_S10-controlled tnp of IS4Bsu1 (data not shown). However, the expression level of the transposase gene in the NBS040 strain seems sufficient not to affect cell growth of B. subtilis while at the same time promoting the transposition of mini-IS.

Some transposons require a host factor(s) for transposition. We showed that in B. subtilis 168, modified IS4Bsu1 isolated from B. subtilis (natto) can transpose, at either high or low frequency, under different conditions (Table 3). These results suggest two possibilities. First, IS4Bsu1 does not require a host factor and depends only on itself. Second, IS4Bsu1 requires a host factor(s) but is able to transpose in B. subtilis 168 because the bacterium supplies this factor(s). There are different conditions that induce mini-IS transposition, and significantly high homology has been shown between the 168 and natto strains. We favour the second possibility because, interestingly, no difference or bias was observed in transposition or target sites under the different conditions tested (Fig. 3, Table 4), suggesting strongly that the difference in transposition frequencies under these conditions is not due to a modification of DNA recognition by transposase per se.

The in vivo regulation of transposition is still poorly understood. Our goal was to explore transposition-inducing conditions and we succeeded in transferring an active IS4Bsu1 transposition system into B. subtilis 168. Our transposition assay system may be a powerful tool for a better understanding of the regulation of cellular transposition.

We thank Mariko Kadota (Musashino University) and Dindo Y. Reyes (Oregon Health and Science University) for helpful discussion and critical reading of the manuscript. We also thank Hiroko Migita, Shuhei Matsuyama, Toshihisa Matsui, Atsushi Kamei, Sachiko Asanuma and Ryobun Santoh for their technical assistance. This work was supported by a Grant-in-Aid for Scientific Research on Priority Area (C) (Genome Biology) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Edited by: M. Hecker