Functional Tn5393-Like Transposon in the R Plasmid pRAS2 from the Fish Pathogen Aeromonas salmonicida subspecies salmonicida Isolated in Norway

Abstract

Tn5393c containing strA-strB was identified as part of R plasmid pRAS2 from the fish pathogen Aeromonas salmonicida subsp. salmonicida. This is the first time an intact and active transposon in the Tn5393 family has been reported in an ecological niche other than an agricultural habitat.

Resistance to streptomycin is often due to genes encoding aminoglycoside-modifying enzymes, and these genes are widely distributed in the environment. The linked strA-strB genes encode two different phosphotransferases, which are both required for high resistance to streptomycin (4). This gene pair has been reported in both pathogenic and environmental bacteria from humans, animals, and plants. In human- and animal-associated bacteria, the strA-strB genes are usually borne on small, nonconjugative, broad-host-range plasmids (18). The strA-strB genes in plant-pathogenic bacteria are generally associated with Tn5393-like transposons which typically reside on large (30- to 220-kbp) conjugative plasmids (17).

In the present study, we examined the occurrence of a Tn5393-like transposon on a 48-kb conjugative R plasmid, pRAS2, from the fish-pathogenic bacterium Aeromonas salmonicida subsp. salmonicida strain 1682/92. In addition to streptomycin resistance, this plasmid also conferred resistance to sulfonamides (sulII) and tetracycline (Tet 31; unpublished data) (Fig. 1). We designated the transposon Tn5393c to distinguish it from transposon Tn5393 from Erwinia amylovora, Tn5393a from Pseudomonas syringae pv. syringae, and Tn5393b from Xanthomonas campestris pv. vesicatoria, which are all essentially identical except for the presence of IS elements (Fig. 2) (17).

FIG. 1.

Map of plasmid pRAS2 from A. salmonicida subsp. salmonicida strain 1682/92. Tn5393c is shown in black, while the resistance genes are shown as dark grey arrowed boxes; the determinants Tet 31, which consists of the two genes tetR(31) and tetA(31), and sulII are located outside Tn5393c, while strA-strB is located inside the transposon. P represents restriction sites for PstI, and the brace between the PstI sites refers to the fragment cloned in pUC19. The heavy line below the map indicates the 3-kb region of pRAS2 amplified by PCR.

FIG. 2.

Genetic map of transposon Tn5393c from pRAS2. IS elements IS6100 and IS1133 are found included in the transposon from the plant-pathogenic bacteria X. campestris pv. vesicatoria (Tn5393b) and E. amylovora (Tn5393), respectively. The direction of transcription is indicated by the arrowed boxes representing the genes. The gene positions are based on the published sequence of Tn5393 (3, 17). The heavy lines below the genetic map indicate the three different DNA sources used to prepare the sequence in AF262622. Sequence 1 is derived from the PCR product indicated in Fig. 1, sequence 2 is derived after transposition to pGEM-3Zf(+), and sequence 3 is derived by using the 6.0-kb PstI fragment cloned in pUC19 as the template. IR_L and IR_R, left and right inverted repeats, respectively.

The strA-strB genes were first detected by hybridization using the ECL direct nucleic acid labeling and detection systems (Amersham International plc, Little Chalfont, Buckinghamshire, England) on plasmid DNA transferred to nylon membranes by Southern blotting (13). A 538-bp PCR product from the linked strA-strB genes was used as a probe (16). By using a previously made restriction map of pRAS2 (data not published), the strA-strB genes' position could be located. This map was also used when a 6.0-kb PstI fragment containing the right half of the transposon was cloned in pUC19 prior to the initial sequencing (Fig. 1).

By conjugating pRAS2 into Escherichia coli DH5 which already contained the vector pGEM-3Zf(+), a method successfully used with Tn5393 (3), the functionality of the transposon could be observed. Tn5393c was able to transpose using pGEM-3Zf(+) as a recipient replicon, and insertions were observed at three different positions in the vector (Table 1), every time associated with a 5-bp direct repeat. No target specificity was obvious regarding Tn5393c; the flanking regions in general were characterized by long AT tracts, but the data are limited.

TABLE 1.

Comparison of insertion sites after insertion of Tn5393c

Plasmida	Insertion sitesb
Sequences outside IRL
pRAS2	GGCAATAAAAAACCGGCCGATGCCGGTTTTTTTGTGCCTACTGAAA
pGEM-3Zf(+), 1	ATTTTTCTAAATACATTCAAATATGTATCCGCTCATGAGACAATAA
pGEM-3Zf(+), 2	TCTATTCTTTTGATTTATAAGGGATTTTGCCGATTTCGGCCTATTG
pGEM-3Zf(+), 3	CACTCAACCCTATCTCGGTCTATTCTTTTGATTTATAAGGGATTTT
Sequences outside IRR
pRAS2	TGAAAGGTGCCGTCTGACTGAAAAACTATGCCATACCGATCAAATA
pGEM-3Zf(+), 1	AATAACCCTGATAAATGCTTCAATAATATTGAAAAAGGAAGAGTAT
pGEM-3Zf(+), 2	TATTGGTTAAAAAATGAGCTGATTTAACAAAAATTTAACGCGAATT
pGEM-3Zf(+), 3	ATTTTGCCGATTTCGGCCTATTGGTTAAAAAATGAGCTGATTTAAC

^aTn5393 is inserted in the opposite direction of the numerical order of the pGEM-3Zf(+) vector sequence, and the strand shown is complementary to the strand in the GenBank sequence. The numbers 1, 2, and 3 refers to three different insertion sites in the vector. IR_L, and IR_R, left and right inverted repeats, respectively. 

^bThe 5-bp direct repeats are in boldface, and the nucleotides involved in the stem-loop structure in pRAS2 are underlined. 

A PCR with a sulII-specific primer and a primer located adjacent to the left inverted repeat of Tn5393c was used to amplify the region of pRAS2 flanking the left side of the transposon (Fig. 1). The PCR product of approximately 3.0 kb was generated by the Taq Polymerase Core Kit (Qiagen GmbH, Hilden, Germany) in a thermal cycler (Hybaid; Perkin-Elmer, Foster City, Calif.). A Qiaprep spin miniprep (Qiagen GmbH) isolation of pRAS2 was used as a template. In pRAS2, a stem-loop structure of 11-bp inverted repeats separated by 4 bp is present 14 bp prior to the transposon (Table 1). The 5-bp direct repeat is also present in pRAS2.

DNA analyses were performed using the BigDye Terminator Ready Reaction Kit (Applied Biosystems, Perkin-Elmer) with an ABI Prism 377 automated sequencing machine (Perkin-Elmer). Three template sources were used (Fig. 2). Sequence analyses of Tn5393c revealed almost 100% identity to Tn5393 on plasmid pEa34 from E. amylovora (GenBank accession number M96392) (3). The strA-strB genes, the resolvase gene tnpR, and the central recombination site res were identical to the corresponding sequence in Tn5393 from pEa34. The transposase gene tnpA from Tn5393c differed by six bp from Tn5393; three CGs in Tn5393 were GCs in the corresponding sequence of Tn5393c. These three 2-bp conversions, located at nucleotides 299 and 300, 1064 and 1065, and 1362 and 1363 of Tn5393, lead to the change of four amino acids in the transposase protein. However, no IS elements were present in the intergenic region between tnpR and strA; nor was the 3-bp direct repeat (TAG) associated with IS1133 identified. Due to the lack of IS elements, the exact size of Tn5393c is only 5,470 bp, compared to the 6,705 bp of Tn5393, which contains IS1133.

The IS elements in the variants of the Tn5393 family are involved in the expression of strA-strB. Tn5393 (IS1133) yielded a streptomycin MIC of 500 to 1,000 μg/ml in E. amylovora, Tn5393a (without IS) yielded a MIC of 75 μg/ml in P. syringae pv. syringae, and Tn5393b (IS6100) yielded a MIC of 250 μg/ml in X. campestris pv. vesicatoria (17). When no IS element is present in the transposon, a promoter sequence within res expresses the genes along with tnpR as one operon (17). IS1133 provides a promoter for transcription of the resistance genes (3), as does IS6100, although in this case the nature of the promoter sequence remains unknown (17). However, IS6100 in X. campestris is inserted within the tnpR gene, inactivating this gene, whose product represses transcription by binding to res. Hence, the increased MIC associated with IS6100 might be due to the combination of these two factors, a cessation of repression and the active function of the promoter in the IS. In addition, the exact same promoter sequence might function at different levels in different organisms, and the IS's influence on the MICs cannot be compared across the species (17). The MICs of streptomycin, determined by a standard microdilution method (10), for A. salmonicida subsp. salmonicida 1682/92 and the corresponding E. coli DH5 transconjugant illustrate this effect, as they were 1,024 and 64 μg/ml, respectively. MICs of sulfonamides and tetracycline were not reduced in the E. coli transconjugant. The absence of expressed streptomycin resistance in E. coli transformants is also reported with Tn5393-containing R plasmids originating from plant-pathogenic bacteria (14).

In several human- and animal-associated bacteria with strA-strB, there are remnants of the Tn5393 sequence in the regions flanking the determinant (6, 12, 15), suggesting that the strA-strB gene pair occurring on small, nonconjugative plasmids originates from Tn5393 (18). The finding of Tn5393c adds more evidence to this theory by confirming that the intact transposon is distributed beyond the plant and soil habitat. Tn5393a and Tn5393c, the variants of the transposon containing no IS elements, probably represent the ancestors of Tn5393 and Tn5393b.

The presence of antibiotic resistance genes in bacteria is thought to be closely related to the use of the specific antibiotic in the bacterial environment (5). Large amounts of antibiotics like oxytetracycline, sulfonamides-trimethoprim, and quinolones have been used therapeutically to cure salmonids with furunculosis caused by A. salmonicida subsp. salmonicida in Norwegian fish farming (7), and the finding of bacteria with resistance to any of these compounds is not surprising (8, 11). Streptomycin, however, has never been used in the sea to treat disease in cultured fish in Norway. Other fish farming countries, including Japan, Canada, and Scotland, have not used streptomycin to cure outbreaks of diseases like furunculosis in seawater either (2, 9), yet these countries also report streptomycin resistance in fish-pathogenic bacteria (1, 9, 19). The streptomycin determinants conferring resistance in these cases have not been reported. As observed in this study, with Tn5393c integrated in pRAS2, the use of either tetracycline or sulfonamides will, of course, be selective for the entire transposon as well, but the question of why and from where Tn5393c originally disseminated to pRAS2 remains unanswered.

The finding of Tn5393c in A. salmonicida subsp. salmonicida illustrates once more how antibiotic resistance determinants and transposable elements might intermingle between distantly related habitats and how available the common microbial gene pool is.

Nucleotide sequence accession number.

The 5,670-bp GenBank sequence consists of the 5,470-bp transposon Tn5393c and 100 bp of pRAS2 on each side. The GenBank accession number for the 5,670-bp sequence is AF262622.