|Home | About | Journals | Submit | Contact Us | Français|
The aim of this study was to characterize a novel conjugative transposon Tn6009 composed of a Tn916 linked to a Staphylococcus aureus mer operon in representative Gram-positive and Gram-negative bacteria isolated in Nigeria and Portugal.
Eighty-three Gram-positive and 34 Gram-negative bacteria were screened for the presence of the Tn6009 using DNA–DNA hybridization, PCR, hybridization of PCR products, sequencing and mating experiments by established procedures.
Forty-three oral and 23 urine Gram-negative and Gram-positive isolates carried the Tn6009. Sequencing was performed to verify the direct linkage between the mer resistance genes and the tet(M) gene. A Nigerian Klebsiella pneumoniae, isolated from a urinary tract infection patient, and one commensal isolate from each of the other Tn6009-positive genera, Serratia liquefaciens, Pseudomonas sp., Enterococcus sp. and Streptococcus sp. isolated from the oral and urine samples of healthy Portuguese children, were able to act as donors and conjugally transfer the Tn6009 to the Enterococcus faecalis JH2-2 recipient, resulting in tetracycline- and mercury-resistant E. faecalis transconjugants.
This study reports a novel non-composite conjugative transposon Tn6009 containing a Tn916 element linked to an S. aureus mer operon carrying genes coding for inorganic mercury resistance (merA), an organic mercury resistance (merB), a regulatory protein (merR) and a mercury transporter (merT). This transposon was identified in 66 isolates from two Gram-positive and three Gram-negative genera and is the first transposon in the Tn916 family to carry the Gram-positive mer genes directly linked to the tet(M) gene.
The tet(M) gene codes for a ribosomal protection protein, which confers tetracycline, doxycycline and minocycline resistance. The tet(M) gene is usually associated with conjugative transposons (CTns) and has been found naturally in 50 different Gram-positive and 22 different Gram-negative genera.1,2 Tn916 was described in 1980 and is the first member of the Tn916–Tn1545 family of transposons that normally carry a tet(M) gene.3 Tn916 (18 kb) is the smallest CTn of this family and was from an Enterococcus faecalis that was isolated from a patient in Michigan,3 while a second CTn from the family, Tn1545 (25.3 kb), was isolated from Streptococcus pneumoniae at around the same time.4 Tn1545 carried tet(M), erm(B), coding for an rRNA methylase that confers resistance to macrolides, lincosamides and streptogramin B, and an aphA-3 gene, coding for kanamycin resistance.4,5 These CTns are usually self-transmissible elements and are associated with both chromosomes and/or plasmids.1
Many Tn916–Tn1545-like transposons have since been characterized from a variety of bacteria.5–7 Most of the Tn916–Tn1545 transposon family carry the tet(M), or tet(M) and erm(B), or tet(M), erm(B) and aphA-3 genes. However, recently, Lancaster et al.8 identified a unique Tn916-like transposon from Streptococcus intermedius, where the tet(M) structural gene had been replaced by the highly related tet(S) gene, which also codes for a ribosomal protection protein and confers resistance to tetracycline, minocycline and doxycycline. More recently, the mef(A) and msr(D) genes, both coding for different macrolide transporters and conferring resistance to macrolides, have been identified downstream of the tet(M) gene in Tn2009 and Tn2010.9,10
In this study, we describe a novel transposon of the Tn916–Tn1545 family, Tn6009, a non-composite transposon containing a Tn916 element directly linked to a Staphylococcus aureus mer operon with genes conferring inorganic mercury resistance (merA), an organic mercury resistance (merB), a repressor (merR) and a mercury transporter (merT). This new CTn was identified in 13 (38%) of the Gram-negative bacteria and 53 (64%) Gram-positive isolates examined. Tn6009 included the 24 orfs of the Tn916 with the Tn916 orf24 linked to a unique 37 bp sequence followed by the staphylococcal merB, merA, merT and merR genes.
In this study, we examined 117 bacteria that had been previously identified as tet(M) and Gram-positive merA positive11 (authors' unpublished observations) (Table 1). Eighteen genotypically distinct Klebsiella pneumoniae isolates were isolated from different Nigerian patients with urinary tract infections, from 2002 to 2003, and previously characterized.12 The 70 oral and 29 urine bacteria were isolated from 1997 to 2000 from different healthy Portuguese children who participated in a randomized controlled trial designed to assess the safety of amalgams (Table 1).13,14 Mercury resistance (Hgr) was determined by growth on brain heart infusion (BHI) agar (Difco Laboratories, Sparks, MD, USA), supplemented with 100 µM mercury chloride as described previously.11 The mercury-resistant E. faecalis CH116, carrying a 65 kb composite transposon including the tet(M) and mer genes,15 and Tn916-positive E. faecalis DSI63 (GenBank accession no. NC_006372) were used as controls in the PCR assays. The mercury- and tetracycline-susceptible (HgsTcs) E. faecalis JH2-2, resistant to fusidic acid, rifampicin and streptomycin (Fusr Rifr Strr), was used as a recipient in the mating experiments and as a negative control in the PCR assay as described previously.16,17
GeneScreenPlus membrane (NEN Research, Boston, MA, USA) and/or Graph Transfer Membrane (GE Nylon, Minnetonka, MN, USA) were used for the DNA–DNA hybridization of Southern blots, DNA dot blots and/or PCR dot blots and were hybridized using 32P-labelled probes as described previously.16,17 All primers used in this study are listed in Table 2.
The primer pairs used for various PCR amplifications are listed in Table 2. DNA preparation and electrophoresis of PCR products were carried out by established procedures.18 The 24 Tn916 orfs were confirmed using overlapping PCR assays, DNA–DNA hybridization of the PCR products and/or by sequencing (Table 2). The PCR conditions used were as follows: the initial denaturation step was for 3 min at 96°C, amplification was 35 cycles of 30 s at 96°C, 1 min at 55°C and 2 min at 72°C, with an extension for 10 min at 72°C. The PCR products were visualized on 1.0% agarose gels stained in ethidium bromide as described previously.16 The Gram-positive mercury resistance genes (merR, orf2, orf3, orf4, merT, merA, merB and orf1) were identified by PCR assays, and DNA–DNA hybridization of PCR products using specific internal 32P-labelled probes, and verified by sequencing as described previously.16,17
The initial PCR assays were designed to sequence the right and left flanking regions of the Tn916 using a combination of Tn916 primers and arbitrary mer primers as described previously.19 Separate PCR assays were performed using the following pairs of primers: 916-23F and MRAR, 916-22F and MRAR, and 916-22R and MERB1F (Table 2). The cycling parameters were: initial denaturation at 96°C for 3 min, amplification was 35 cycles of 30 s at 96°C, 1 min at 50°C and 2 min at 72°C, with an extension for 10 min at 72°C. The PCR products were visualized on 1.0% agarose gels, and Southern blot transfer was done as described previously.18 Southern blot hybridization was done using 32P-labelled probes of 916-24F and MERB1R (Table 2). The band that hybridized with the probes was excised, and the gel purified and sequenced as described previously.16
Five isolates, K. pneumoniae Kpn 19, Serratia liquefaciens 142, Pseudomonas sp. 613, Streptococcus sp. 7 and Enterococcus sp. 5, were used for conjugation experiments. The recipient was E. faecalis JH2-2, chromosomally resistant to streptomycin (500 mg/L), rifampicin (25 mg/L), nalidixic acid (25 mg/L) and polymyxin B, (100 mg/L) as described previously.16 E. faecalis transconjugants were selected on BHI agar (Difco Laboratories, Division of Becton–Dickinson and Co.) supplemented with tetracycline 10 mg/L and either rifampicin 25 mg/L or polymyxin B 100 mg/L (for K. pneumoniae conjugation experiments), and then verified as E. faecalis by streaking out on CHROMagar™ Orientation (DRG International, Inc., Mountainside, NJ, USA), and verification of drug resistances of the E. faecalis JH2-2 transconjugants to streptomycin (500 mg/L) and/or nalidixic acid (25 mg/L), was done as previously described.16 Transfer of the mer genes was determined by growth on mercury plates (100 µM) and verified by DNA–DNA hybridization and PCR assays. Mating experiments were repeated and frequencies were averaged. For confirmation of a physical linkage between the mer resistance genes and the Tn916 int gene, overlapping PCR assays were done.
All PCR products used for sequence analysis were purified using a Microcon Ultracel YM-100 filter (Millipore Corp., Bedford, MA, USA) following the manufacturer's instructions. The PCR products were sequenced directly or were cloned into the pCR®T7/NT- TOPO® vector (Invitrogen, CA, USA) according to the manufacturer's instructions and the plasmid inserts sequenced using the T7 promoter forward and reverse primers. PCR amplicons or plasmid templates were sequenced bidirectionally at the University of Washington, Biochemistry Sequencing Facility as described previously.16 Homology searches were performed with BLAST and ORF finder tools (http://www.ncbi.nlm.nih.gov). The searches were performed at both the DNA and protein levels. Multiple sequences were aligned with Clustal W (http://www.ebi.ac.uk/clustalw/index.html), and sequences were assembled using Vector NTI Advance 10™ (Invitrogen).
The element was designated Tn6009 by the transposon nomenclature centre (http://www.ucl.ac.uk/eastman/tn/; 24 February 2008, date last accessed). A total of 17 767 bp between Tn916 orf6 and the merT gene was sequenced (Figure 1) and assigned the GenBank accession no. EU239355. A second 1889 bp region, downstream of Tn916 tet(M) and including the orf5, xis and int genes, was also sequenced and given GenBank accession no. EU399632.
The presence of Gram-positive merA genes in 14 Gram-negative bacteria has previously been reported.11 Seven of these isolates also carried the tet(M) gene and we hypothesized that there could be a linkage between the merA and tet(M) genes. To verify this hypothesis, we used PCR assays with one primer within the Tn916 orf22–orf24 region paired with a variety of primers, in both orientations, from merB and merA genes (Table 2). The PCR assay using primer 916-22R, at the very start of Tn916 orf22 gene, and the MERB1F, within the merB gene, gave a reproducible PCR product of 1820 bp, which hybridized with internal probes from Tn916 orf24 and merB. In contrast, the control E. faecalis DS16, which carries Tn916, and E. faecalis CH116, which carries the mer operon, were negative for this assay (data not shown). By sequencing upstream of the Tn916 orf24 gene, a 37 bp region was identified that linked the Tn916 orf24- to the orf1-like gene previously identified in the staphylococcal mer operon (Figure 1).20 This 37 bp region had a 54% G + C, no comparable sequences were identified in GenBank and no tandem repeats or palindromes were found,using the Internet-based analysis tools (http://tandem.bu.edu/trf/trf.basic.submit.html, http://bioweb.pasteur.fr/docs/EMBOSS/palindrome.html; 24 February 2008, date last accessed).
A 21 bp probe (916-MR INT) within the 37 bp region was used to screen the remaining isolates and controls E. faecalis DS16 and E. faecalis CH116. Neither control strain was positive, while 53 (64%) of the Gram-positive isolates, including two genera, and 13 (38%) of the Gram-negative isolates, including three genera, were positive for this 37 bp region using DNA–DNA hybridization and PCR assays (Table 1). Of the nine Gram-negative genera examined, isolates from three genera, Klebsiella, Serratia and Pseudomonas, as well as multiple isolates from both the Gram-positive Enterococcus and Streptococcus genera, were positive (Table 1).
One isolate from each of the five positive genera was selected for further study. The five isolates carried the 24 orfs of the Tn916 linked to the staphylococcal merA, merB, merT and merR genes (data not shown). The mer genes were merR→merT→merA→merB→orf1 for all five isolates (Figure 1).
The Tn6009 from K. pneumoniae Kpn 19 was further characterized. A 17 767 bp region spanning 19 orfs (merT through Tn916 orf6) was sequenced. A second 1889 bp region, downstream of Tn916 tet(M) and including the orf5, xis and int genes, was also sequenced (Figure 1). Homology at the DNA and amino acid levels was compared with sequences from GenBank accession no. NC_006372 from the E. faecalis strain DS16. Amino acid identity ranged from 97% for the 39 amino acid orf24 to 100% for the orf6, int and xis genes, while the DNA and amino acid homology was 100% for the merT, merA and merB genes when compared with GenBank accession no. L29436 from S. aureus plasmid pI258 (Table 3).
To verify that the Tn6009 could be conjugally transferred, three Gram-negative and two Gram-positive isolates, K. pneumoniae 19, S. liquefaciens 142, Pseudomonas sp. 613, Streptococcus sp. 7 and Enterococcus sp. 5, were selected as donors for mating with the E. faecalis JH2-2 recipient. All five donors were able to transfer the Tn6009 at frequencies ranging between 10−6 and 10−8 per recipient. The highest frequency of transfer (10−6) occurred with the Enterococcus sp. donor. The Streptococcus sp. and K. pneumoniae donors had transfer frequencies 10-fold less (10−7), while the S. liquefaciens and Pseudomonas sp. donors transferred at frequencies of 10−8 (Table 4). To verify that the Tn6009 transferred as a single unit, the unique 37 bp sequence was demonstrated, and overlapping PCR assays were performed that included the region between the mer and tet(M) genes. All of the PCR assays produced the expected PCR product that was indistinguishable in size from the original donors. In contrast, the E. faecalis DSI6, which carries Tn916 but not the mer genes, produced no PCR products when one primer was within any of the mer genes and the other primer within any of the 24 orfs of Tn916. The data indicate that the complete Tn6009 was present in the transconjugants (data not shown).
This study identified a new transposon, Tn6009, which is the first transposon to have a direct linkage between mer genes and the tet(M) gene. However, previously a composite transposon, Tn5385, was identified, which carried both the mer and tet(M) genes on separate elements that could potentially transfer independently.15 Tn6009 was identified in Gram-negative bacteria from Nigeria and Portugal and Gram-positive bacteria from Portuguese subjects. We hypothesize that the carriage of Tn6009 may explain why the Gram-positive mer genes were previously identified in Gram-negative mercury-resistant isolates from Portuguese subjects.11
In Tn6009, the mer operon was linked to a complete Tn916 by a unique 37 bp region, while the mer orf1 gene had a deletion of the terminal 268 bp sequences directly upstream of the 37 bp unique region. This suggests that the deletion may have occurred during the integration of the mer operon while the Tn6009 transposon was being formed. The Tn6009 mer genes had the same genetic organization as the staphylococcal mer operon from plasmids pI258 and pMS97 (Figure 1, GenBank accession no. AB179623) and from SCC composite islands.20,21 The Tn6009 was mobile and all five donors were able to transfer the complete transposon into the E. faecalis recipient, resulting in transconjugants that were resistant to mercury, due to the action of the merA gene, and tetracycline, due to the presence of the tet(M) gene. No other phenotypic traits were transferred, though it is possible that an unexpressed antibiotic resistance gene could have been transferred, as previously described in Bacterioides.22
We found that 44% of the K. pneumoniae, 62% of the other Gram-negative and 30% to 50% of the Gram-positive isolates characterized in the laboratory carried both mobile Gram-positive merA and tet(M) genes, but did not carry the unique 37 bp sequence or the Tn6009 element. This implies that there may be other conjugative elements carrying the Gram-positive merA gene, which may or may not be linked to the tet(M).
The Tn6009 was identified in Gram-positive bacteria from Portugal isolated in 1997 and was more common among the Gram-positive bacteria tested, though 33% of the Nigerian K. pneumoniae isolated in 2002–03 carried the Tn6009. At this time, we do not know when or where Tn6009 first appeared, or if it has spread beyond Nigeria and Portugal. Only examination of more isolates from various geographic locations and time periods will give us this information. A much better understanding of the pressures that shape the creation of these new CTns has been formed, and of why some of these elements are able to spread quickly while others stay more localized. Perhaps most importantly, resources are needed to monitor these changes and to develop ways of reducing the pressures, which will allow the continued acquisition and modification of the Tn916 family of transposons.
This study was supported in part by grant U01 DE-1189 and contract N01 DE-72623 from the National Institute of Dental and Craniofacial Research of the National Institutes of Health.
None to declare.
We are grateful to Dr Adam Roberts for helpful suggestions.