INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

Insertion sequences (ISs) are a diverse group of small, mobile genetic elements that can insert into target DNA molecules with various degrees of site specificity. In prokaryotes, more than 500 such elements are now known (for a recent review, see Mahillon and Chandler [15]), and more examples of these intrinsically cryptic DNA segments continue to be identified as the output of known bacterial genomic sequences increases. Although IS elements are typically compact, encoding only functions that are necessary for transposition, they have long been recognized for their ability to influence gene expression, either by directly inactivating genes upon insertion or by providing promoters that increase the transcription of downstream genes (10). In addition, IS elements can effect chromosome rearrangements, including insertions, deletions, inversions, and duplications (17). Consequently, the mobility of such elements can have a profound impact on the stability of IS-containing genomic regions as well as influence the phenotype, or the antigen repertoire, of an organism.

Despite the considerable DNA sequence and functional diversity among ISs, comparisons of multiple elements have identified several characteristic features that are shared by most ISs. Most IS elements are between 800 and 2,500 bp long and contain two important structural components, a gene encoding a transposase and inverted repeat (IR) structures at the termini. Transposases are basic, DNA-binding proteins (typically 250 to 400 amino acids long) that target the terminal IR sequences during the transposition reaction. The IR sequences are usually 10 to 40 bp long, although a few IS elements that lack these terminal structures are known (15). During transposition of most IS elements, the two strands of the target DNA are asymmetrically cleaved, resulting in the generation of short direct repeats (DRs) of the target sequence (one abutting each terminus of the IS) at the conclusion of the insertion reaction (10). Although the extent of such target site duplication varies among elements (usually 2 to 14 bp), the length of the DR is generally fixed for each IS (15).

Based upon differences within the conserved features of IS elements, 17 IS families have been recognized on the basis of one or more of the following criteria: IS open reading frame (ORF) organization, conserved signature motifs among transposases, similarity of IR sequences, and length of target site duplications (15). Each family is named after the prototype member of the group. During the characterization of the Mycoplasma fermentans genomic region that encodes the potent MALP-2 immunomodulin, a previously unrecognized IS, designated IS1630, was identified (3). The putative IS1630 transposase has significant homology to IS30 transposase, but initial characterization of the insertion site and target site duplication suggested that IS1630 may have properties that not only distinguish it from the 15 characterized IS30 family members but also had not been observed for any IS element. Thus, the original cloned copy of IS1630 was flanked by 23-bp DRs, a duplication larger than any previously reported target site. Furthermore, the duplicated sequence was a perfect IR sequence that was derived from a candidate transcription terminator for the malp gene. In contrast, the IS30 prototype creates only 2-bp duplications, and although it exhibits a preference for insertion into sequences that contain a degree of dyad symmetry (18), there is not an absolute requirement for target sequences to conform to a perfect palindrome.

Prompted by the possibility that IS1630 may exhibit novel target site specificity and generate unusually long duplications of perfect IR sequences, we investigated the element further. This report describes the characterization of nine IS1630 copies from the genomes of two M. fermentans strains. Analyses of the flanking regions support our supposition that the processes of target site recognition and duplication by IS1630 are unlike those described for any other IS. In addition, a copy of the previously identified M. fermentans IS element ISMi1 (13) was cloned. This IS3 family member is unrelated to IS1630. Characterization of the reading frames within this copy of ISMi1 led to a reevaluation of the coding potential for this element and the proposal of a model for the translation of a functional transposase.

During the course of this study, two IS elements of the IS4 family, IS1634 in Mycoplasma mycoides subsp. mycoides (29) and IS1548 in Mycobacterium smegmatis (20), have been reported to generate large duplications of variable length, although neither element appears to have the target site recognition and duplication properties described here for the IS30 family member IS1630.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Bacteria, plasmids, and growth conditions. M. fermentans PG18 (clone 39) and M. fermentans II-29/1 were propogated in Hayflick medium and GBF-3 medium respectively, as described previously (3). Escherichia coli DH10B (Gibco BRL, Grand Island, N.Y.) and strains containing pZero2 or pZero2.1 (Invitrogen, Carlsbad, Calif.) and its derivatives were grown at 37°C in Luria-Bertani medium supplemented with 50 µg of kanamycin per ml.

DNA preparation and hybridization methods. Genomic DNAs from M. fermentans PG18 and M. fermentans II-29/1 were prepared as described previously (3). For Southern hybridization, genomic DNA was digested to completion with either EcoRI or HindIII, resolved by electrophoresis in 0.7% (wt/vol) agarose gels, transferred to positively charged nylon membranes (Boehringer Mannheim, Indianapolis, Ind.), and hybridized with an oligonucleotide probe labeled at the 3' end with digoxigenin (DIG) (primer 1, 5' GAA GGA ACA TCA ATA TTA GGA TA). Hybridization and washing steps were carried out at 50°C with standard buffer solutions described in the guide provided by the membrane manufacturer (Boehringer). Hybridizing bands were detected by nonradioactive detection methods as described previously (31). Colony hybridization of transformants with DIG-labeled oligonucleotide primer 1 was carried out following the recommendations of the manufacturer of the nonradioactive labeling and detection system (Boehringer).

Cloning of IS1630 copies. BglII-, HindIII-, or XbaI-digested chromosomal DNA from M. fermentans PG18 or BglII-digested total DNA from M. fermentans II-29/1 was ligated to BamHI-, HindIII-, or XbaI-digested cloning vector pZero 2 or pZero2.1 under conditions recommended by the vector supplier (Invitrogen). Following transformation into competent E. coli DH10B cells (Gibco BRL), kanamycin-resistant transformants were screened by standard colony hybridization methods with DIG-labeled oligonucleotide primer 1 (see above). Plasmid preparations (24) were analyzed for differently sized inserts that hybridized with probe 6 (5' ATT AGG TTA TTC AAG AAC AAC TA). Plasmid clones containing individual IS1630 copies were generated by use of the following ligations: IS1630A (4-kb NheI fragment from strain PG18 in the XbaI site of pZero2.1); IS1630B, IS1630C, and IS1630D (BglII fragments from PG18 in the BamHI site of pZero2); IS1630E (XbaI fragment from PG18 in the XbaI site of pZero2); IS1630F (HindIII fragment from PG18 in the HindIII site of pZero2.1); and IS1630G, IS1630H, and IS1630I (BglII fragments from strain II-29/1 in pZero2).

DNA sequencing and computer analysis. The nucleotide sequence for each of the cloned IS1630 copies was determined with six oligonucleotide primers that were located within the IS element. These were primer 1 (see above), primer 2 (5' TAG TTG CTC AAA GTA AGT ATT CAA), primer 3 (5' TGT AAG TTT ACA CCT AGA AAT TTG), primer 4 (5' TAA TCC ATT ATC CTG AGT TAT AG), primer 5 (5' CCA AAT ATT TTA GGG AGG GCT A), and primer 6 (see above). Flanking DNA sequences were determined with the outwardly facing primers 1 and 2, the cloning vector primers SP6 and M13F, and sequence-generated custom oligonucleotide primers. All oligonucleotides used in this study were synthesized on a model 3948 Nucleic Acid Synthesis and Purification System (Applied Biosystems, Inc., Foster City, Calif.), and all DNA sequencing was performed with Taq Dye terminators and a Prism 377 automated DNA sequencer (Applied Biosystems, Inc.). Both of these services were carried out at the University of Missouri Molecular Biology Program DNA Core Facility. DNA and protein sequences were analyzed by use of the GCG software package (Genetics Computer Group, Madison, Wis.).

PCR amplification and chromosomal linkage analysis of M. fermentans strains. Oligonucleotide primers 5 and 6 were used to amplify a 254-bp portion of IS1630 from M. fermentans genomic DNA preparations (kindly provided by S.-C. Lo, Armed Forces Institute of Pathology, Washington, D.C.). These included DNAs from strains K7, MT-2, M39A, M70B, SK5, SK6, and Incognitus. The DNA template (10 ng) was used in a standard PCR consisting of 35 cycles at 94°C for 1 min, 50°C for 1 min, and 72°C for 1 min. The resulting PCR amplicons were purified and analyzed by agarose gel electrophoresis.

Inverse PCR methods. Inverse PCR was used to isolate the DNA sequence downstream of IS1630B from M. fermentans PG18. Briefly, approximately 500 ng of genomic DNA was digested with HindIII, purified, and self-ligated by use of procedures described previously (3). Samples of the ligation were amplified with the Expand Long Template PCR System under conditions recommended by the product supplier. With primers B-inv1 (5' CAT CAA TGC AAA ATT CAA GTG TG) and B-inv2 (5' CTA AGG AAA CTA TTT CAA TAT CAT), a 3.5-kb fragment was generated; this size was expected based on prior Southern hybridization analysis with DIG-labeled primer B-inv1 as the probe. The resulting amplicon was purified and sequenced directly by primer walking.

Nucleotide sequence accession numbers. The sequences reported in this paper have been deposited in GenBank under accession no. AF100324 (IS1630A), AF179373 (IS1630B), AF179374 (IS1630C), AF179375 (IS1630D), AF179376 (IS1630E), AF179377 (IS1630F), AF179378 (IS1630G), AF179379 (IS1630H), and AF179380 (IS1630I).

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

Identification of an IS30-type IS element in M. fermentans. During a comparison of the malp chromosomal regions of M. fermentans PG18 and II-29/1, a restriction fragment length polymorphism that was detected for several restriction enzymes proved to be due to the presence (in strain PG18) or absence of a previously unrecognized IS-like element (3). Designated IS1630 (by the Plasmid Reference Center, Stanford University School of Medicine, Stanford, Calif.), this insertion element is 1,377 bp long and is bounded by 27-bp IRs (Fig. 1). IS1630 has an A+T content of 74%, which is close to the 72% A+T composition of the 15-kb genomic region within which this IS element was first identified. A single ORF encoding either 366 or 387 amino acid residues (depending on which of two candidate AUG start codons is used) occupies most of this compactly organized element. Analysis of the codon usage showed that 11 of the 14 Trp residues in the predicted product are encoded by UGA codons, indicating that this IS element is indigenous to mycoplasmas (16) and has not been transferred recently to M. fermentans from a nonmycoplasmal source.

View larger version (46K): [in this window] [in a new window]   FIG. 1.   Nucleotide sequence and selected features of IS1630. The nucleotide sequence of IS1630A and flanking regions is shown, with the nucleotide numbers at left. Nucleotide 1 is the first nucleotide of the left IR of the IS. The deduced amino acid sequence encoded by the single ORF is shown in single-letter code below the DNA sequence, and each of the three acidic residues that comprise the highly conserved active-site motif (DDE) is enclosed within a box. The two candidate methionine start codons are underlined. The limits of IS1630 are indicated by square brackets, and the IR sequences at the left (IR-L) and right (IR-R) termini are highlighted by solid black lines. IR sequences within IS1630 are indicated by opposing arrows. The DR sequences that flank IS1630A are highlighted with asterisks below the corresponding sequence. The nucleotide sequence is presented in the orientation opposite that in GenBank AF100324.

IS1630 copy number in two M. fermentans strains. The data presented above suggested that IS1630 could generate large DR sequences upon transposition. To determine whether this function was a consistent feature of the element, multiple independent insertion events needed to be characterized. The absence of gene transfer methodology for M. fermentans and the presence of 11 UGA Trp codons in the transposase coding sequence prevented the application of standard genetic techniques for recovering multiple insertions in either mycoplasmas or E. coli, respectively. As an alternative, Southern analysis was used to determine whether multiple copies of IS1630 (and, therefore, IS1630 insertion sites) were present in the genomes of two M. fermentans strains. DNA from clonal isolates of strains PG18 and II-29/1 was restricted and subjected to Southern transfer and hybridization analysis with an IS1630-derived oligonucleotide as a probe. The oligonucleotide sequence was selected from a region of IS1630 which does not exhibit detectable homology to other IS elements. Under high-stringency hybridization and washing conditions, multiple hybridizing bands were observed in both restriction digests (EcoRI and HindIII) from each strain (Fig. 2). Both restriction digests of PG18 DNA yielded seven hybridizing bands, suggesting that there are at least seven copies of IS1630 in the genome of this strain. DNA from strain II-29/1 produced six hybridizing bands in the EcoRI digest and five in the HindIII reaction, although the intensity of hybridization of an approximately 3-kb HindIII fragment may indicate the presence of two hybridizing sequences in this band. Therefore, there are probably a minimum of six copies of IS1630 in strain II-29/1.

View larger version (63K): [in this window] [in a new window]   FIG. 2.   Identification of multiple copies of IS1630 in M. fermentans strains. Genomic DNA from M. fermentans PG18 (lanes 2 and 4) or II-29/1 (lanes 3 and 5) was digested with EcoRI (lanes 2 and 3) or HindIII (lanes 4 and 5), transferred to a nylon membrane, and hybridized with DIG-labeled oligonucleotide probe 5 (see Materials and Methods). Lane 1 contains DIG-labeled lambda HindIII markers (Boehringer), the sizes of which are indicated in kilobase pairs.

Comparisons among individual IS1630 copies. To characterize IS1630 and its associated DR sequences further, multiple copies were cloned in E. coli and identified by colony hybridization. In this way, five additional IS1630-containing fragments were cloned from strain PG18 and three were isolated from strain II-29/1. Sequence analysis of the nine cloned copies indicated that each was flanked by a unique nucleotide sequence, demonstrating that each clone contained an independent copy of the element together with its insertion site (Fig. 3). Each copy of IS1630 was given a specific designation based on its unique insertion site (IS1630A to IS1630F for those from strain PG18; IS1630G to IS1630I for those from strain II-29/1), with the original cloned copy having the IS1630A designation. With the exception of the truncated element, IS1630F, which contains only the 5' 748 bp of the element, each of the other copies was similar in length to IS1630A.

View larger version (24K): [in this window] [in a new window]   FIG. 3.   Location of indels within IS1630 variants. Each cloned copy of IS1630 is shown as a thick horizontal line. IS1630A and IS1630C lack nucleotide (nt) indels. The locations of indels within the other IS1630 variants are shown by a nucleotide number above the horizontal line (numbered according to the IS1630A sequence). The nature of the indel is shown below the line. The positions of two candidate methionine start codons (M), the stop codon (stop), and the residues that comprise the DDE motif are shown below the horizontal line representing IS1630A. The 10 nucleotides immediately downstream of each IS copy are shown at the right, except for the truncated IS1630F, which lacks the distal portion of IS1630.

Location of cloned copies of IS1630. The nucleotide sequences that flank each of the cloned IS1630 copies were determined and compared to those in GenBank. The gene organization of these IS1630-linked regions is shown in Fig. 4, and the closest homologs of each gene product (as revealed by the FastA program) are presented in Table 1. Although approximately half of the gene products encoded by the flanking genes had recognizable homologs in the current database, eight of the gene products lacked identifiable counterparts from other bacterial genomes, including the fully sequenced chromosomes of Mycoplasma genitalium (8) and Mycoplasma pneumoniae (11). One striking feature resulting from this analysis of IS location is that in each case for which it was possible to determine location, IS1630 was inserted in an intergenic region close to a stop codon of a neighboring gene. It was not possible to readily determine the genes that flank IS1630I, as the clone harboring this copy of the element contains only a few nucleotides between the IR sequences and the restriction site used for cloning. As indicated above, IS1630I has a copy of ISMi1 inserted at nucleotide 43. 

View larger version (20K): [in this window] [in a new window]   FIG. 4.   Gene organization of cloned IS1630 flanking regions. The location and direction of ORFs flanking each cloned IS1630 element are shown by open arrows. IS1630 is represented by a grey arrow (rectangle in the case of the truncated IS1630F), and ISMi1 is shown as a solid black arrow. The number of base pairs between the IS1630 termini and the adjacent ORFs is shown between the arrows. Two tRNA genes downstream of IS1630C are also shown (T). When an ORF could not be given an unambiguous gene assignment, even if it exhibited homology to genes of unknown function in other bacteria, a designation that indicated to which IS1630 copy the ORF was linked was given (e.g. orfG1 and orfG2 flank IS1630G).

IS1630 generates large target site duplications. Members of the IS30 family of IS elements typically generate 2- or 3-bp duplications of the target site upon insertion (15, 18). The original IS1630 copy was flanked by a 23-bp DR, suggesting that IS1630 can create longer duplications. To determine whether or not this was a general rule for IS1630 insertion, the nucleotide sequences abutting the left and right IR sequences were analyzed for the six additional IS copies for which both flanking sequences were available. For five copies, perfect DRs ranging from 19 to 26 bp flanked the IS1630 element. The exception, IS1630D, lacked an apparent target site duplication (a possible reason for this finding is discussed below). These data are consistent with the hypothesis that IS1630 transposition creates large duplications but cannot rule out the formal possibility that IS1630 is inserted between tandem DRs.

View larger version (23K): [in this window] [in a new window]   FIG. 5.   Determination of insertion site occupancy in two M. fermentans strains by PCR. Amplicons generated by PCR on genomic DNA from M. fermentans PG18 (lanes 1, 4, 7, 11, and 14) or II-29/1 (lanes 2, 5, 8, 12, and 15) or water-containing negative controls (lanes 3, 6, 9, 13, and 16) were separated by agarose gel electrophoresis. Each series of three reactions contained a pair of opposing oligonucleotide primers that were derived from sequences flanking a specific IS1630 copy: IS1630A (lanes 1 to 3), IS1630B (lanes 4 to 6), IS1630H (lanes 7 to 9), IS1630C (lanes 11 to 13), and IS1630E (lanes 14 to 16). For any primer pair, a size difference of 1.4 kb between strains indicates that a site is occupied in one strain and empty in the other. When a site is occupied in both strains, the resulting PCR products are equal in size. Lanes 10 and 17 contain a 1-kb ladder (Promega Corporation, Madison, Wis.). The positions of the 1-kb (open triangle) and 3-kb (solid triangle) size markers are indicated.

IS1630 is inserted into IR sequences that resemble transcription terminators. Comparisons of the DR sequences failed to identify a well-conserved consensus sequence for target site recognition. However, each DR sequence is itself an IR sequence. Furthermore, the two different sequences that flank the termini of IS1630D and the sequence that flanks the single IS1630 terminal IR of IS1630F are also IR sequences. Therefore, IS1630 has apparent specificity for IR sequences. The finding that IS1630D is not flanked by long DR sequences but abuts two different IR sequences raises the possibility that IS1630D is a hybrid IS1630 element that has arisen by recombination between two similarly oriented IS elements that were each inserted into a unique IR. In this regard, it should be noted that the distal 150-bp portion of IS1630D is the most divergent from that present in IS1630A (11 nucleotide substitutions), whereas the proximal 1,200-bp portion of this element is the most similar to IS1630A (10 nucleotide substitutions). Such asymmetry in the distribution of nucleotide substitutions is not seen with any of the other IS1630 variants and is consistent with the notion that IS1630D arose by recombination between two IS elements.

View larger version (17K): [in this window] [in a new window]   FIG. 6.   Features of IS1630 insertion sites. Predicted RNA secondary structures are shown for putative transcription terminators encoded by DNA sequences that are IS1630 target sites. The C-terminal amino acid (single-letter code: K, E or N) and stop codon (*) are shown at the left of each secondary structure. Three empty sites [(E)] were sequenced to confirm the stem-loop sequence. For orfE1, orfG2, and orfC1, the structures shown were derived by deleting the sequence of IS1630 and one copy of the DR. For orfD2 and orfF1, the appropriate IS1630 flanking sequences are not present to allow empty target sites to be deduced. The sites of insertion are shown by the solid triangles. The sequences duplicated as DRs are the DNA sequences that encode the stem-loop structures shown above the pairs of triangles. nt, nucleotides.

IS1630 is widely distributed among M. fermentans isolates. To address the strain distribution of IS1630 further, primers 5 and 6 were used in PCR to amplify a 254-bp region of the IS. Amplicons of the expected sizes were obtained for each of the seven M. fermentans isolates tested (data not shown). These included MT-2, SK5, SK6, Incognitus, K7, M39A, and M70B. The cloning of IS1630 variants from strains PG18 and II-29/1 raises the number of IS1630-containing M. fermentans strains to nine. The wide distribution of IS1630 among M. fermentans strains suggests that sequences within IS1630 are potentially useful diagnostic tools for the PCR-based detection of the human infectious agent M. fermentans. IS1630 is not present in the complete genome sequences of M. genitalium (8) and M. pneumoniae (11), but its possible presence in numerous other human mycoplasmas, both pathogens and commensals, requires determination.

Further analysis of ISMi1 from M. fermentans II-29/1. As reported above, IS1630I from M. fermentans II-29/1 contains a copy of the IS3-type element ISMi1. Although the sequence of this ISMi1 copy is >98% identical to that described for the original IS (13), the sequence differences warranted a reevaluation of the coding potential of this element. The sequence of the original copy contained two potential ORFs separated by a 541-bp untranslated sequence. The two ORFs, designated ORF1 and ORF2, encoded 143 amino acid residues and 109 amino acid residues, respectively. The ORF2 product had a short region (57 amino acid residues) of homology to transposases of the IS3 group (13).

View larger version (40K): [in this window] [in a new window]   FIG. 7.   Features of ISMi1. (A) Alignment of the deduced amino acid sequences encoded by ORF2* from the copy of ISMi1 inserted into IS1630I and ORF2 from IS150. Identical amino acid residues are highlighted with a black background. The numbers at the right indicate the positions of the amino acid residues for the respective ORFs (residue 1 for ORF2* is the first amino acid for the 1 reading frame after the stop codon for ORF1). A black vertical arrow indicates the originally proposed (13) position of the methionine start site for ORF2, and an open vertical arrow indicates the start site for the C-terminal region encoded by ORF2*, which is absent in ORF2. The residues comprising the DDE motif are indicated by asterisks. Gaps in the alignment are shown by dashes. (B) Proposed model for a putative frameshifting window in ISMi1 and comparison of the revised (upper cartoon) and originally proposed (lower cartoon) ORF arrangements in this element. A stem-loop structure for the ISMi1 mRNA region in which ribosomal frameshifting is proposed to occur is shown, together with the 0 and 1 reading frames. The ORF1 stop codon (UAA) is highlighted with asterisks in the secondary structure model and in the 0-frame sequence. The same three nucleotides are underlined in the 1 reading frame, in which they are components encoding consecutive Asn codons. In the schematic diagram showing the features of ISMi1, ORF1 is shown as a black arrow and ORF2 and ORF2* are represented by open arrows. The left (IR-L) and right (IR-R) IRs are shown as grey vertical bars at the IS termini. The location of the proposed frameshifting window is indicated by a triangle. The horizontal lines indicate untranslated regions. (C) Features of the frameshift window that has been established for IS150, highlighted as in panel B. The ORF1 stop codon in this case is a UGA codon. The features of IS150 are shown in the schematic diagram, following the conventions used in panel B.

Concluding remarks. Prior to this study, the sequences and features of five IS elements from mycoplasmas had been reported. Four of these, IS1138 of Mycoplasma pulmonis (2), IS1221 of Mycoplasma hyorhinis and Mycoplasma hyopneumoniae (32), IS1296 of M. mycoides (9), and ISMi1 (also known as ISLE) of M. fermentans (13), are members of the IS3 family, the largest of the 17 major groupings of IS elements that were recently reviewed (15). The only non-IS3 family IS elements to be described so far in mycoplasmas are the IS4-type element IS1634 of M. mycoides (29) and the IS30 family member IS1630, described in this report. Although IS1630 possesses a putative transposase with hallmark signature sequence motifs for the group and has overall similarity in organization to other IS30 members, IS1630 has expanded the spectrum of insertion site specificity and the extent of target site duplication. Three IS elements that create long DRs have now been described, and while those that have been identified so far for IS1630 are shorter than those for the IS4-type elements, the symmetry of the IRs within the duplication has not been observed previously.