|Home | About | Journals | Submit | Contact Us | Français|
The bacterial transposon Tn917 inserts preferentially in the terminus region of some members of the Firmicutes. To determine what molecular process was being targeted by the element, we analyzed Tn917 target site selection in Bacillus subtilis. We find that Tn917 insertions accumulate around the central terminators, terI and terII, in wild-type cells with or without the SPβ lysogen. Highly focused targeting around terI and terII requires the trans-acting termination protein RTP, but it is unaffected in strains compromised in dimer resolution or chromosome translocation. This work indicates that Tn917 is sensitive to differences in DNA replication termination between the Firmicutes.
Certain transposons are known to target features associated with DNA metabolism, and these elements have the potential to offer greater insight into these host processes (6). The bacterial transposon Tn917 was originally isolated from Enterococcus faecalis and has been used as an insertion mutagen in this and other gram-positive bacteria (5, 28, 32). However, Tn917 has been shown to have an extreme regional preference for insertion into the terminus region in E. faecalis, and the molecular mechanism responsible for this bias is unknown (10). Multiple processing events associated with chromosome duplication occur in the terminus region, a portion of the chromosome we define here as equidistant from the origin of DNA replication in circular genomes. In some bacteria, a system exists that actively terminates DNA replication forks at specific sites within the terminus region, called ter sites, through the use of a trans-acting protein called Tus in Escherichia coli and RTP in Bacillus subtilis (8, 9, 18). Another processing event that occurs in the terminus region involves the resolution of dimer chromosomes at a site called dif (2). The resolution of dimer chromosome is usually catalyzed by two tyrosine recombinases called XerC and XerD in E. coli and RipX and CodV in B. subtilis.
To better understand processing events in bacterial chromosomes, we investigated Tn917 target site selection in the model low-G+C gram-positive bacterium, B. subtilis, where replication and recombination are well studied. We were specifically interested in knowing if Tn917 targeted DNA replication termination or dimer resolution as suggested previously (10). There are multiple examples where elements have been suggested to recognize these molecular processes as targets for transposition in E. coli (20, 23, 29). In the case of Tn917, it is of special interest to understand targeting because this behavior differs even between closely related species; Tn917 transposition preferentially occurs in the terminus region of Enterococcus faecalis and Streptococcus equi, but this behavior is not found in Listeria monocytogenes and Streptococcus suis (see references 10 and 28 and see below).
Tn917 insertions were collected in B. subtilis using plasmid pTV1-OK in the strain CU1065 (W168 trpC2 SPβ−) background using a procedure that prevented the isolation and sequencing of the same transposition event (11). The temperature-sensitive plasmid pTV1-OK imparts resistance to kanamycin and contains the erythromycin resistance-encoding transposon Tn917 (11). To generate transposants, single colonies of purified pTV1-OK transformants (12) were used to inoculate individual test tubes containing LB liquid medium with kanamycin (10 μg/ml) plus erythromycin at a sublethal concentration (1 ng/ml), a level previously shown to induce the erm and transposase genes of Tn917 (30). Cultures were incubated at 30°C overnight and then plated on LB medium containing erythromycin (1 μg/ml) and incubated at 42°C overnight. The latter incubation step was repeated to ensure loss of the transposon delivery plasmid. We determined the position and orientation of individual transposition events in the chromosome using arbitrary PCR analysis and the sequence of the B. subtilis 168 chromosome version AL009126.1 as described previously (10, 11, 14). This analysis indicated that transposition events occur preferentially around the first ter sites encountered by DNA replication forks in the chromosome, terI and terII (Fig. (Fig.1A);1A); 30% (25/82) of the transposition events occurred within 15 kb of these sites even though this region comprises less than 1% of the chromosome (Fig. (Fig.1C).1C). Previously it was shown that Tn917 transposition showed a strong bias for the gltA gene in screens for auxotrophs, a gene known to be close to the region where DNA replication terminated (25, 31, 32) and now known to be ~3.6 kb from terI-terII (14). However, our work here is the first indication that transposition actually occurred around the central terI-terII terminators.
To further confirm the Tn917 preference for the region around the terI-terII sites, we analyzed targeting in B. subtilis strain JH642 containing the lysogen SPβ. In this strain the terI-terII sites move 134 kb relative to oriC from the position found in B. subtilis strain CU1065 (15). We found that Tn917 insertions continued to have a preference for the terI-terII sites in the lysogen strain where 17% (4/23) of the insertions still occurred within 15 kb of these sites even though this region comprises less than 1% of the chromosome (data not shown). This supported the idea that targeting is not dependent on the relative position in the chromosome but is instead a sequence or process directly associated with this particular region of the chromosome.
To determine if targeting required active termination of DNA replication, we examined Tn917 transposition in an Δrtp strain. Active termination is not an essential process in bacteria, and the only phenotype associated with an Δrtp allele is dependent on the inactivation of other systems (17). We created a CU1065 Δrtp::cat strain using long-flanking-homology PCR analysis (19). The oligonucleotide primers JEP158 (5′-GGGTAACTAGCCTCGCCGGTCCACGATATTAAAGACTGATAGTCC-3′) and JEP159 (5′-CCGGCATCAGCAAATTTGGCGG-3′) were used to amplify the region 5′ to the deletion; JEP137 (5′-AATGCTTCGGCCAGCTTCTTCAGG-3′) and JEP138 (5′-CTTGATAATAAGGGTAACTATTGCCTTTAATAGAAACAAACACC-3′) were used to amplify the region 3′ to the deletion. The primers and plasmids used for amplification of the antibiotic resistance cassettes have been described previously (3). Deletion of the rtp gene in the B. subtilis chromosome was confirmed by PCR analysis.
We found that there was still a general preference for Tn917 transposition across the terminus region in the Δrtp background (e.g., in both the rtp+ and Δrtp backgrounds about 40% of the insertions occurred within a 200-kb window centered around the terI-terII sites) (Fig. 1C and D). However, the extreme preference for the region around the terI-terII sites was lost in the Δrtp background, and only 6% (5/83) of the transposition events occurred within 15 kb of the terI-terII sites (Fig. (Fig.1B).1B). Our data indicate that within the terminus region, while insertions occur at a greater-than-expected frequency within 15 kb of the central terI-terII sites in the wild-type background (P < 0.001; χ2 statistic), this was not true in the Δrtp strain (P = 0.22; χ2 statistic). These data are consistent with a model where Tn917 targets the region where DNA replication terminates in the chromosome. While there is no requirement for active termination of DNA replication, the RTP-mediated process likely focuses Tn917 insertions around the central terI-terII sites.
In E. faecalis, Tn917 insertions occur with a strong grouping where 65% of the insertions occurred in a 200-kb region (1,450 to 1,650 kb) centered around the predicted natural position of replication termination in this organism as indicated by the skew of the chromosome (10, 13, 16). This position also correlates with the dif site used to resolve dimer chromosomes in E. faecalis (1,550,523 bp), and 23% of the insertions occurred within 15 kb of the predicted dif site (a region which constitutes ~1% of the chromosome) (13, 16). While Tn917 did not appear to target the dif site in B. subtilis (Fig. (Fig.1C),1C), we wanted to decisively rule out any role of dimer resolution in Tn917 targeting. The RipX and CodV proteins are involved in dif recombination in B. subtilis, but the ripX gene product is known to play the essential role in this process. Therefore, we monitored transposition in an otherwise isogenic ΔripX B. subtilis strain which is deficient in chromosome dimer resolution (26). A CU1065 ΔripX::cat strain was constructed using transformation with chromosomal DNA from strain PAL422 (ΔripX::cat) (27). The distribution of Tn917 insertions indicates that there is still a preference for transposition in the terminus region (Fig. (Fig.2A).2A). In addition, Tn917 insertions still preferentially occurred around terI-terII within the terminus region in the ΔripX strain; 22% (17/77) of the insertions occurred within 15 kb of terI-terII sites in the ΔripX background (Fig. 2A and C). Similar to the result found with the wild-type strain, we found that insertions occur at a greater-than-expected frequency within 15 kb of the central terI-terII sites within the region shown in Fig. Fig.2C2C in the ΔripX background (P < 0.001; χ2 statistic). This confirms that dimer resolution is not responsible for the attraction of Tn917 insertion for the terminus region in B. subtilis.
We also determined if the DNA translocation protein SpoIIIE played any role in targeting Tn917 insertions to the terminus region. SpoIIIE monitors signals in the bacterial chromosome that convene in the terminus region. In E. coli, the chromosome dimer resolution proteins require a partner protein, FtsK, for completing recombination (2, 4). The FtsK and SpoIIIE proteins are both able to monitor DNA sequences in the chromosome and translocate DNA in one direction relative to the chromosomal dif site (reviewed in reference 1). While the role of the SpoIIIE protein in actively growing cells is unclear, we were still interested in this gene product because Tn917 inserts at a region in the chromosome where the sequences recognized by SpoIIIE converge. A CU1065 spoIIIE::spc strain was constructed using transformation with chromosomal DNA from strain KPL708 (spoIIIE::spc) (24). After examining the distribution of insertions, we found that there was still a significant bias for Tn917 insertion within the terminus region where 20% (16/77) of the insertions fell within 15 kb of the terI-terII sites in the ΔspoIIIE background (P < 0.001; χ2 statistic) (Fig. 2B and D). Our results indicate that neither the B. subtilis dimer resolution system nor the translocation function of SpoIIIE is responsible for attracting Tn917 events to the terminus region of the chromosome. B. subtilis has two poorly understood proteins with homology to DNA translocases like SpoIIIE and FtsK called YtpT and YtpS (26). The YtpT protein is not required for dif recombination as monitored in a plasmid-based assay, and any role for the YtpT and YtpS proteins in Tn917 targeting was not investigated here. The result with the ΔripX and ΔspoIIIE strains also indicates that a direct interaction with the RipX and SpoIIIE proteins does not direct Tn917 transposition into the terminus region.
In summary, we can now confirm that Tn917 transposition events are attracted to the place where DNA replication forks are expected to terminate through the action of the RTP protein in the B. subtilis chromosome in a process that is unaffected by the dimer resolution and DNA translocation systems (Fig. (Fig.11 and and2).2). Other transposons have been shown to target features of DNA replication termination in bacteria (23, 29). In Tn7 target site selection, at least two molecular signals are used to recognize active DNA replication, a gapped DNA structure and an interaction with the β-clamp processivity factor (21). Presumably these features become more available when DNA replication terminates. Interestingly, it was also noted that the Tn917 transposase has a sequence that resembles a protein motif used by a variety of proteins to interact with the processivity factor (21). It seems possible that in both Tn7 and Tn917 transposition, the β-clamp processivity factor may provide part of the signal for identifying insertion sites and that the β-clamp may become available when replication is terminated by either passive or active means. Further research will be needed to confirm that a sequence within the Tn917 transposase interacts with the processivity factor and if this interaction is important for recognizing replication termination and perhaps other replication targets (e.g., DNA repair signals and the replication of mobile DNA elements, as in the case of Tn7 ).
Tn917 appears to be sensitive to an unknown difference in either how DNA replication is terminated or how termination events are processed within the Firmicutes. While Tn917 insertions are focused around the predicted point of replication termination in E. faecalis, Tn917 targeting did not show this bias in Listeria monocytogenes (10). A similar discrepancy was found within the closely related Streptococcus equi and Streptococcus suis species. Tn917 insertions did not occur with any obvious bias in the S. suis genome; however, in S. equi, 60% of the Tn917 insertions occurred in a 15-kb region (28). By comparing the genes found in this region of the chromosome with the DNA sequence from S. equi, we can report that Tn917 also targets the terminus region in S. equi (personal observation), a region where dimer chromosomes are likely to be resolved via an unconventional system (difSL) found in the Streptococcus and Lactococcus genera (16). This is a region where replication termination is likely to occur in S. equi based on the skew of the genome (16).
In E. faecalis and in S. equi, the grouping of Tn917 insertions was surprisingly strong despite the absence of an active replication termination system (23% and 60% of the insertions, respectively, fell in a region around dif that made up about 1% of the chromosome [10, 28]); in B. subtilis, insertions were found only to focus tightly in a very small region in the presence of active termination via RTP (30% of the insertions in a region around terI-terII that comprised about 1% of the chromosome) (compare Fig. 1C and D). This could suggest that an accessory termination system may be acting in E. faecalis and S. equi to actively terminate DNA replication. While dimer resolution systems have been suggested to be capable of terminating DNA replication (13), work in E. coli using 2-D electrophoresis suggests that only a very small percentage of DNA replication forks actually stall or slow around the dif site (7). It is formally possible that dimer resolution systems may show an altered ability to stall or slow replication forks in different species of bacteria.
Our results firmly establish that Tn917 is capable of recognizing features of replication termination when selecting where to insert in the chromosome. Further research will be needed to determine if accessory proteins in some members of the Firmicutes alter the ability of Tn917 to preferentially target where DNA replication terminates. Alternatively, processing of replication forks following replication termination may differ in some fundamental way within the Firmicutes.
We thank Nancy Craig (Howard Hughes Medical Institute, Johns Hopkins Medical School), David Sherratt (Oxford University), and Alan Grossman (MIT) for providing reagents. We thank Brian Swingle (Cornell University) for comments on the manuscript.
This work was funded by grant GM069508 from the National Institutes of Health.
Published ahead of print on 9 October 2009.