INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Filamentous soil bacteria of the genus Streptomyces contain a linear chromosome about 8 Mb in size (24). In addition, they frequently carry more than one linear plasmid (20). These linear molecules can recombine with each other to form hybrid structures. For example, SCP1, a linear plasmid of 350 kb (21, 22, 31), interacts with the chromosome of Streptomyces coelicolor A3(2) and generates various SCP1-chromosome hybrid structures (10, 13, 23). During the analysis of these structures, we found several insertion sequences (ISs) near the target site for SCP1 integration in strain M145.

ISs, the simplest transposable elements, usually encode one protein required for transposition (8, 27). They are found at many sites in host genomes and can cause insertional inactivation of genes. In addition, ISs contribute to genome rearrangements such as translocation, duplication, inversion, and deletion. Since the discovery of IS110 from S. coelicolor A3(2) (5), several transposable elements have been isolated from Streptomyces species, and some of them were involved in genome rearrangements. For example, IS1373 was found in Streptomyces lividans as a direct repeat flanking the mercury resistance genes in AUD2, which was a unit for large-scale DNA amplification (39). Tn4811 (6) is present near both ends of the linear chromosome of S. lividans 66 as well as near the right end of the linear plasmid SLP2. This fact led Lin et al. (24) to the finding that the chromosome of S. lividans is linear.

In this study, we identified a total of five copies of four IS elements in a 6.5-kb DNA segment of the chromosome of S. coelicolor M145. Among the four IS elements, three (IS468, IS469, and IS470) are new and the other is IS466 (18). The transposition activity of IS466 was experimentally verified. The properties of these IS elements are described on the basis of their nucleotide sequences. Possible functions of the IS cluster are discussed in relation to the linear structure of the S. coelicolor A3(2) chromosome and its dynamic rearrangements.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains, cosmids, and plasmids. Streptomyces strains, cosmids, and plasmids used in this study are listed in Table 1. All of the S. coelicolor A3(2) strains are derivatives of the wild-type strain 1147 (11), which carries both SCP1 and the 30-kb circular plasmid SCP2 (25). Strain M145 (SCP1 SCP2) is a reference strain of S. coelicolor A3(2) now being used for the genome project. Escherichia coli SURE was used as recipient for the cosmid library. E. coli XL1-Blue was used for cloning and sequencing of DNA fragments in pUC19 and phage M13mp18/19. The cosmid library of S. coelicolor A634 was constructed using a Gigapack packaging kit and Supercos1 vector (Stratagene, La Jolla, Calif.) as described previously (32). pUN121, which was used for cloning of a transposed copy of IS466, was described by Nilsson et al. (29).

DNA isolation and sequencing. Streptomyces strains were cultured in YEME medium and total DNAs were isolated as described by Hopwood et al. (16). E. coli strains were cultured in Luria-Bertani medium, and the cosmid and plasmid DNAs were isolated as described by Sambrook et al. (34). The 8.6-kb KpnI-EcoRI fragment of cosmid E46 was digested with appropriate restriction endonucleases and cloned into M13mp18/19. Nucleotide sequencing was carried out by the dideoxy termination method using a dye terminator cycle sequencing kit (Amersham Pharmacia Biotech, Uppsala, Sweden) and the ABI-373 sequencing system (PE Biosystems, Foster City, Calif.). Genetyx-Mac 7.3 (Software Development, Tokyo, Japan) was used for analysis of the sequence data.

Hybridization experiments. DNA fragments were separated by agarose gel electrophoresis and transferred to nylon membranes. Hybridization was carried out using the Boehringer Mannheim DIG system overnight at 70°C in standard buffer according to the supplier's protocol. After hybridization, washing was done twice for 5 min each in 2× wash solution at room temperature and then twice for 15 min each in 0.1× wash solution at 70°C.

Nucleotide sequence accession number. The sequence data for the 8.6-kb KpnI-EcoRI fragment have been submitted to the DDBJ/EMBL/GenBank databases under accession no. AB032065.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Finding of large IS cluster. We preliminarily analyzed the SCP1-chromosome hybrid structures in S. coelicolor A3(2) strains by using the cloned SCP1 end fragment as a probe (23). For more precise analysis, we used the ordered cosmid libraries for both SCP1 (31) and the chromosome of S. coelicolor M145 (32), and a newly constructed cosmid library for an SCP1-integrated strain A634. In strain A634, SCP1 is integrated into AseI fragment E in the center of the chromosome, and the target site for integration is located on cosmid E46 from strain M145 (M. Yamasaki et al., unpublished results).

View larger version (9K): [in this window] [in a new window]   FIG. 1.   Organization of the IS cluster in S. coelicolor M145. Based on the nucleotide sequence of the KpnI-EcoRI fragment (8,632 bp) of cosmid E46, the alignment of ORF X, the five IS elements, and dagA is depicted. The nucleotide sequence of the right EcoRI-BamHI fragment was reported by Buttner et al. (2). IS elements are indicated by boxes, and ORFs are marked by open arrows. Restriction sites for BamHI, EcoRI, EcoRV, KpnI, SalI, SphI, and XhoI are also shown. Bars below the map indicate the DNA probes which were used as probes for hybridization to determine the copy numbers of the IS elements.

View larger version (34K): [in this window] [in a new window]   FIG. 2.   IR and DR sequences of the five IS elements. Ninety nucleotides around the left and right ends of IS468A, IS468B, IS469, IS466A, and IS470 are shown, from which the IR and DR sequences of each IS element were deduced. The IR and DR sequences are indicated by closed arrows and shaded boxes, respectively, on the side of the reading strand. The ORFs coding an unknown protein, transposases, and agarase are shown by open arrows.

Characterization of IS468A and -B. IS468A and -B are identical in nucleotide sequence and are tandemly aligned with a common DR sequence (CTCAG) between them (Fig. 2). IS468 is 1,097 bp long (nt 1849 to 2945; nt 2951 to 4047) and contains a 9-bp perfect IR at both ends. The GC content of IS468 is 62.2%, which is much lower than the average for Streptomyces DNA (70 to 74% [4]). This suggests that IS468A and -B were transferred relatively recently to S. coelicolor A3(2) and have not been affected by a pressure for high GC content in Streptomyces species. Frame analysis revealed that IS468A and -B each contain one ORF (ORFA and ORFB, respectively) starting at nt 2071 and 3173 and stopping at nt 2910 and 4012, respectively. Each ORF would encode a 31.5-kDa protein with 279 amino acid residues (aa).

Target sequence for IS468. Hybridization experiments revealed an additional copy of IS468 (IS468C) on the linking cosmid E2, which connects AseI fragments E and H (32). We also detected another copy of IS468 (IS468S) in strain A634, which was localized on an SCP1 molecule integrated into AseI fragment E by analysis of cosmid 2G12. However, IS468S is not present on a freely replicating SCP1 molecule in the wild-type strain 1147. Therefore, we cloned and sequenced the corresponding region of SCP1 in 1147 using cosmid 63 (31) and found there a target sequence (CTGAG) in place of IS468S.

View larger version (26K): [in this window] [in a new window]   FIG. 3.   DR sequences flanking different copies of IS468 and IS466. (A) The DR sequences flanking IS468A, -B, and -C in S. coelicolor M145 and IS468S in A634, and the target sequence on the free SCP1 in 1147. IS468A and -B were subcloned from cosmid E46, IS468C was subcloned from cosmid E2, IS468S was subcloned from cosmid 2G12, and the target sequence was subcloned from cosmid 63. (B) The DR sequences flanking IS466A in S. coelicolor M145 and its transposed copy in S. lividans TK64.

Characterization of IS469. IS469 is 1,627 bp long (nt 4072 to 5698) and contains a 28-bp imperfect (27/28) IR (these sizes may be changed slightly by analysis of other copies). The DR sequence could not be identified due to an overlap with IS466 or rearrangements (Fig. 2). The GC content is 67.9%, slightly lower than the average for Streptomyces DNA. Frame analysis revealed that IS469 contains one ORF starting at nt 4157 and stopping at nt 5662, which encodes a large protein with 501 aa and a molecular mass of 56.1 kDa. The calculated pI for the predicted protein is 10.3.

Characterization of IS466. IS466 was first detected as a repeated sequence that was present in two copies near the agarase gene (dagA) on the S. coelicolor M145 chromosome and also on SCP1 (18). The latter has been located at the inside end of the right terminal IR of SCP1 (10). IS466 is not present in the closely related strain S. lividans 66.

View larger version (106K): [in this window] [in a new window]   FIG. 4.   Alignment of the amino acid sequence of the IS469 transposase with those of IS466, J30-ORF16 (J30-16), IS1001, P. putida tnpA (Ps tnpA), IS21 istA, and IS640 istA. For the last four transposases, only the similar regions are shown. Identical and similar amino acids are shaded in black and gray, respectively. Similar amino acids are grouped as follows: MVLI; AGSTP; QEND; FWY; KRH; C. The conserved D, D, and E residues are indicated by asterisks.

Characterization of IS470. IS470 is 1,008 bp long (nt 8366 to 7359; GC content, 67.8%) and contains a 15-bp imperfect (13/15) IR and a 5-bp flanking DR (CAGTG). IS470 contains one ORF on the complementary strand (nt 8225 to 7398). It encodes a 30.1-kDa protein with 275 aa (calculated pI, 10.0). The protein product of IS470 has significant homologies to the transposases coded by IS493-ORFB in S. lividans (38% identity and 59% similarity [35]) and cosmid 6G9-ORF5 in S. coelicolor A3(2) (35 and 53%; accession no. AL079356).

Copy numbers of IS elements. In many cases, IS elements are found in several copies. To count IS468A and -B separately, total DNAs of various S. coelicolor A3(2) strains were digested with SphI and probed by the SphI-SalI fragment of IS468A (probe 1 in Fig. 1). As shown in Fig. 5A, strain M145, which carries neither SCP1 nor SCP2, gave five hybridizing fragments of 4.6, 4.0, 3.2, 1.1, and 1.0 kb. Among them, the 3.2- and 1.0-kb fragments correspond to IS468B and -A, respectively. On the other hand, the wild-type strain 1147, which carries both SCP1 and SCP2, gave three hybridizing fragments of 4.6, 4.0, and 3.2 kb. These results indicate that the chromosome of the wild-type strain 1147 differs from that of plasmid-free strain M145. This is important to note because the S. coelicolor genome project now in progress uses a cosmid library of strain M145. The result of the SCP1-NF strain 2612 is quite interesting, because none of the five fragments of IS468 was detected. Strain 1984, which carries the hybrid plasmid SCP1'-cysB, also shows a totally different hybridization pattern.

View larger version (87K): [in this window] [in a new window]   FIG. 5.   Southern hybridization analysis of various S. coelicolor A3(2) strains probed by the four IS elements. To determine the copy number of each IS element, total DNAs of S. coelicolor A3(2) strains were digested with the indicated enzymes and hybridized with the indicated DNA probes (Fig. 1): (A) IS468, SphI and probe 1; (B) IS470, SalI and probe 4; (C) IS469, BamHI and probe 2; (D) IS466, BamHI and probe 3. HindIII fragments of DNA were used as size markers.

Transposition of IS466. The temperature-sensitive plasmid pMT664 (30), which carries IS466A, was introduced into S. lividans 66 strain TK64. First, galK mutants were selected as described by Kendall et al. (19) and were expected to include IS466 insertion mutations into the gal operon. Thirty mutants were tested for survival on thiostrepton-containing medium (50 µg/ml) at the restrictive temperature (39°C). Three mutants showed enhanced survival possibly due to the transposition of IS466 into the chromosome, which in turn provided homology to allow plasmid integration. The three strains were cured of the plasmid by culturing at 39°C without thiostrepton and then analyzed.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In this study, we identified a cluster of five copies of four insertion sequences (IS468A, IS468B, IS469, IS466A, and IS470) near the target site for SCP1 integration in S. coelicolor M145 and next to the agarase gene (dagA), a gene which is frequently deleted by chromosomal rearrangements. IS468A and -B are aligned tandemly at the left end of the cluster. The tandem repeat might have been formed by sequential insertions of two copies of IS468 into the target sequence (CTCAG) or deletion of an intervening sequence of a large transposable element flanked by two copies of IS468. IS468S was detected on the SCP1 molecule integrated into the chromosome in strain A634, but not on the free SCP1 molecule in strain 1147. It was revealed that the corresponding region of SCP1 in 1147 contains the target sequence (CTGAG) instead of IS468S. This confirmed that IS468 had transposed into the target sequence and generated IS468S.

It is noteworthy that IS469 and IS466A are quite similar to each other (69.3% homology) and are located in a tail-to-tail orientation. The activity of IS466 was experimentally proved by transposition to S. lividans. The 8-bp flanking duplications in S. coelicolor A3(2) and S. lividans were different, which suggests that the target sequence for IS466 may be fairly random. The sequences flanking the transposed copy of IS466 in S. lividans did not come from the gal operon (1). As high transposition rates have been observed in the transposon Tn5424 constructed from IS466 (17), it is possible that the galK mutation and the IS466 insertion were independent events.

Interaction of SCP1 and the S. coelicolor A3(2) chromosome generates two types of SCP1-chromosome hybrid structures; SCP1-integrated chromosomes and SCP1' plasmids (13). This is reminiscent of the plasmid F and the E. coli chromosome. In E. coli Hfr strains, the F plasmid is integrated into the chromosome by homologous recombination between two insertion sequences, IS2 and IS3, located on both the F plasmid and the chromosome, or by transposition into nonspecific chromosomal sites. Both of IS2 and IS3 are present in at least five copies on the E. coli chromosome (36). Homologous recombinations between these and other IS elements cause various chromosomal rearrangements in E. coli strains.

In S. coelicolor A3(2) strains 2612, A634, A610, and A608, SCP1 is integrated into the chromosome (13, 23). These strains showed a hybridization pattern different from that of the wild-type strain 1147 when probed by the four IS elements. This suggests that the IS cluster may have been involved in the formation of these SCP1-integrated strains. It was reported that in strain 2612, IS466 constitutes the right junction of the integrated copy of SCP1 (10). However, the integration of SCP1 into the S. coelicolor A3(2) chromosome cannot be explained by a single cross-over, because this would split the chromosome. Similar hybridization analysis of IS110 (5) and IS117 (minicircle) (26) done previously indicated some variations in a pedigree of S. coelicolor A3(2) strains.

S. coelicolor 1984 and 2106 carry SCP1-chromosome hybrid plasmids SCP1'-cysB (550 kb) and SCP1'-cysD (1,700 kb), respectively (15, 23). Different from the F' plasmids in E. coli, which are circular, these SCP1' plasmids are linear; at least one of the end fragments was detected by hybridization (23). The hybridization pattern with probe 1 (Fig. 5A) suggests that IS468 was involved in the formation of the hybrid plasmid, SCP1'-cysB in 1984.

The unique IS cluster found in this study might have other functions than the interaction of SCP1 and the S. coelicolor A3(2) chromosome. The Streptomyces chromosome is about 8 Mb in size, or about twice as large as the E. coli chromosome. Comparison of the gene organizations of two genera suggests that the Streptomyces chromosomes were generated by fusion of two E. coli-sized chromosomes (12). The IS cluster found in this study is located near the center of the chromosome. The latest data of the S. coelicolor genome project revealed that more than 20 transposase genes are present near the left end of the M145 chromosome (the right-end region has not yet been analyzed). Among them are one of the remaining copies of IS468 (cosmid J11-ORF27, accession no. AL109949) and the IS element similar to IS466 and IS469 (cosmid J30-ORF16). None of the five copies of IS468 was detected in the SCP1-NF strain 2612, which indicates that the left end region as well as the central region were rearranged. All of the data together suggest that IS elements play an important role in dynamic rearrangements of the S. coelicolor A3(2) chromosome and may have been involved in the fusion of two E. coli-sized chromosomes to generate the present form.