The native Pseudomonas stutzeri strain Q chromosomal integron can capture and express cassette-associated genes

Nicholas V. Coleman and Andrew J. Holmes

School of Molecular and Microbial Biosciences, Building G08, University of Sydney, NSW, 2006, Australia

Correspondence
Andrew J. Holmes
A.Holmes{at}mmb.usyd.edu.au


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The integron-gene cassette system contributes to multiple antibiotic resistance in bacteria and is likely to be of broader evolutionary significance. However, the majority of integron diversity consists of chromosomal integrons (CIs), with mostly unknown phenotypes, which are poorly characterized. A pUC-based reporter plasmid (pUS23) was developed containing a recombination site [aadB 59 base element (59-be)] upstream of promoterless aadB [gentamicin (Gm) resistance] and gfp (green fluorescence) genes, and this construct was used to investigate the recombination and expression activities of the CI in Pseudomonas stutzeri strain Q. Electroporation of pUS23 into P. stutzeri Q gave ampicillin-resistant transformants, which yielded GmR green fluorescent recombinants after plating on Gm medium. Site-specific integration of pUS23 at attI was detected by PCR in 8 % of GmR colonies and the frequency of attI integration was estimated as 2·0x10–8 per P. stutzeri Q(pUS23) cell. RT-PCR confirmed integron-mediated expression of aadB in one recombinant strain (Q23-17) and a promoter (Pc) was localized to the 5' end of the intI gene. The integrated pUS23 and flanking integron DNA were cloned from genomic DNA of strain Q23-17 and sequenced, confirming that site-specific integration of the entire reporter plasmid had occurred at the attI site. An insertion sequence (ISPst5; IS5 family) was discovered in the vector backbone of the reporter plasmid integrated at attI and also in a pUS23 derivative recovered as a plasmid in Escherichia coli JM109. This is the first demonstration that wild-type CIs can capture gene cassettes and express cassette-associated genes.


Abbreviations: Ap, ampicillin; Cm, chloramphenicol; Gm, gentamicin; Tc, tetracycline; 59-be, 59 base element; CI, chromosomal integron; gcORF, gene-cassette-associated ORF; MI, mobilized integron

The GenBank/EMBL/DDBJ accession numbers for the sequences of IS element ISPst5 and of pUS23 reported in this paper are AY894752 and AY894753, respectively.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Integrons and gene cassettes function together as a distinctive genetic system (Fig. 1). Gene cassettes typically consist of a single promoterless gene and an associated recombination site (termed 59 base element, 59-be or attC). The total pool of cassette-associated genes in natural bacterial populations (gene cassette metagenome) is a significant source of genetic novelty (Holmes et al., 2003a; Michael et al., 2004; Stokes et al., 2001), and since individually packaged genes create the possibility of combinatorial genetics, the significance of the gene cassette metagenome is potentially far reaching. However, gene cassettes are dependent on externally supplied functions for both mobilization and expression. These functions are supplied by integrons (Hall & Collis, 1995; Rowe-Magnus & Mazel, 2001). The key features of an integron are a recombination site (attI), an enzyme (integron integrase, IntI) that catalyses site-specific recombination between cassette and integron sites (attIx59-be or 59-bex59-be) and a cassette promoter (Pc). Collectively, these give an integron the capacity to acquire gene cassettes, express gene-cassette-associated ORFs (gcORFs) and rearrange the transcriptional order of gcORFs through repeated integration and excision of gene cassettes.



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 1. Structure of pUS23 reporter plasmid and mechanism of capture by InPstQ. (a) Reporter construct pUS23: bla, {beta}-lactamase (ApR); pUC ori, pUC19 origin of replication; Plac, lac promoter of pUC19; T1T2, rrnBT1T2 transcription terminators; aadB 59-be (attC site) associated with aadB gene; aadB, promoterless aminoglycoside adenylyltransferase (GmR KmR TbR); gfp, promoterless GFP (mutant 3). All restriction sites indicated are unique, except KpnI, and are abbreviated as follows; E, EcoRI; K, KpnI; Sm, SmaI; B, BamHI; X, XbaI; S, SalI; P, PstI; Bg, BglII; A, ApaI; H, HindIII. (b) P. stutzeri Q transformant containing pUS23 and InPstQ. Expression of IntIPstQ integrase leads to site-specific recombination between the aadB 59-be and the attIPstQ site, and integration of the plasmid into InPstQ. Also indicated are 59-bes associated with the native gene cassettes of P. stutzeri Q (BGC001–003) and the integron-associated promoter Pc. The integrase promoter is not shown, but is expected to be located within or near the attIPstQ site. (c) Recombinant strain containing pUS23 integrated at attI. The aadB and gfp genes are expressed from the integron-associated Pc promoter. Note that the wild-type PstQ integron contains 10 gene cassettes, of which only the first three are indicated in this figure.

 
Interpretation of the significance of the gene cassette metagenome is dependent on understanding the distribution and diversity of integrons. Under the selective pressure of intensive antibiotic use, class 1 integrons have used the cassette metagenome to engineer the phenotype of multiple antibiotic resistance (Hall & Collis, 1998; Rowe-Magnus et al., 2001). Examination of genome sequence data and targeted environmental surveys have shown that integrons are diverse and present in ~5 % of bacterial genomes (Nemergut et al., 2004; Nield et al., 2001; Rowe-Magnus et al., 2001). Given the size of the gene cassette reservoir and the relative frequency of integrons, we might expect that the integron-gene cassette system has also engineered phenotypes other than antibiotic resistance. Surprisingly, such evidence is hard to find, suggesting that other integron classes have been less successful in this regard than class 1 integrons. A contributing factor may be that not all integrons have the same opportunities to access the gene cassette metagenome.

Integrons have been observed in a diverse range of genetic elements, including transposons, plasmids, genomic islands, chromosomes and combinations thereof (Boyd et al., 2002; Hall & Collis, 1998; Hochhut et al., 2001; Liebert et al., 1999; Rowe-Magnus et al., 2001; Szczepanowski et al., 2004). The typical cassette array characteristics of those integron classes associated with mobile elements (mobilized integrons, MI) differ from those integrons that are only known from bacterial chromosomes (chromosomal integrons, CI). In comparison to MIs, CIs tend to contain larger cassette arrays, have significantly greater uniformity in 59-bes and predominantly contain cassette-associated genes of unknown function. These differences are likely to reflect different capacities to move to new sources of gene cassettes, but do they also reflect diversity in integron functions? It has been suggested that class 1 integrons have a greater capacity to express cassette-associated genes than other integron classes (Hanau-Bercot et al., 2002) and even within class 1 integrons different versions of Pc are known, which vary significantly in their promoter strength (Collis & Hall, 1995; Levesque et al., 1994). The issue of variation in integron function is particularly significant in the case of CIs since they can form long-term associations with specific bacterial lineages (Gillings et al., 2005; Rowe-Magnus et al., 2003, 2001). Are such integrons fully functional or primarily cassette reservoirs?

Available data are limited, but indicate that CIs share the site-specific recombination functions of class 1 (Collis et al., 2001) and class 3 (Collis et al., 2002a) integrons. The integron integrases from CIs of various species have been expressed in Escherichia coli and shown to catalyse site-specific recombination between attI or 59-be sites (Drouin et al., 2002; Holmes et al., 2003b; Leon & Roy, 2003; Rowe-Magnus et al., 2001). However, changes in cassette arrays within a single strain have not been observed over time in lab cultures (unpublished data) and it is notable that integration, excision or Pc-mediated expression of gcORFs have not yet been reported in a wild-type CI. A particularly significant issue is the extent to which cassette-associated genes in the very large cassette arrays of some CIs can be expressed. We describe here an approach for addressing these issues. A model gene cassette was constructed by inserting 59-be and promoterless reporter genes into a pUC19 backbone. When transformed into cells containing integrons this reporter cassette (pUS23) can provide evidence for both integron integrase and Pc functionality. We used the reporter pUS23 to investigate the CI in Pseudomonas stutzeri strain Q (InPstQ) (Holmes et al., 2003b) and report here the first evidence of IntI-mediated gene capture at attI and expression from Pc in a wild-type CI.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and growth conditions.
Bacterial strains are described in Table 1. All bacteria were grown aerobically in LB medium (Sambrook & Russell, 2001) at 37 °C. Gentamicin (Gm) was added at 10 µg ml–1 for E. coli strains and at 3 µg ml–1 for Pseudomonas strains unless indicated otherwise. Chloramphenicol (Cm) was added at 25 µg ml–1 and tetracycline (Tc) at 10 µg ml–1 for E. coli strains. Ampicillin (Ap) was added at 100 µg ml–1 for both genera. Minimal inhibitory concentrations (MICs) were determined in broth cultures inoculated at an OD600 of approximately 0·05 and grown for 16 h.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids

 
Microscopy.
Cells from late-exponential-phase broth cultures (OD600=1·0–1·5) were washed in sucrose buffer (0·3 M sucrose, 1 mM MgCl2, 1 mM HEPES, pH 7), concentrated 10-fold and 5 µl samples were mounted on pads of 1 % agarose in the same buffer. An Olympus BX60 microscope with a Chroma filter set 41018 (Chroma Technologies) was used for fluorescence microscopy. Images were captured with a SenSys digital CCD camera (Photometrics) and processed using V for Windows software (Digital Optics). An exposure time of 50 ms was used for all experiments.

General DNA techniques.
Plasmids and oligonucleotides used are described in Table 1 and Table 2, respectively. Plasmid DNA extraction, restriction digestion and ligation were done by standard methods (Sambrook & Russell, 2001). Qiaquick columns (Qiagen) were used to purify DNA for ligations and sequencing. Molecular biology enzymes were from Genesearch (New England Biolabs). DNA sequencing of plasmids and PCR products was done by the dye-terminator method on an ABI 3730 machine (SUPAMAC facility, University of Sydney). Genomic DNA was extracted by a CTAB (N-cetyl-N,N,N-trimethylammonium bromide) method as follows. Cells from 50 ml broth cultures were resuspended in 5 ml STE buffer (1 M NaCl, 10 mM Tris, 1 mM EDTA), then CTAB (0·5 ml, 10 % w/v) was added and incubated at 65 °C for 30 min. Chloroform/isoamyl alcohol (24 : 1 ratio, 5 ml) was added, mixed thoroughly and the lysate incubated on ice for 30 min before centrifugation (20 000 g, 10 min, 4 °C). The DNA-containing supernatant was further purified with phenol/chloroform and ethanol (Sambrook & Russell, 2001). Both E. coli and Pseudomonas strains were transformed by electroporation (2·5 kV, 25 µF, 200 m{Omega}, 0·2 cm gap cuvettes), essentially using standard methods (Sambrook & Russell, 2001), except that Pseudomonas cells were harvested in late exponential phase (OD600=1·0) and washed and resuspended in sucrose buffer (see above). Recovery after electroporation was in LB medium (1 ml) for 1 h (E. coli) or 2 h (Pseudomonas) at 37 °C before cells were plated on antibiotic media.


View this table:
[in this window]
[in a new window]
 
Table 2. Oligonucleotides

 
PCR.
PCR was done in 25 µl reactions containing Pfu polymerase (1 U), dNTPs (0·2 mM) and primers (1 µM each) in a buffer containing Tris/HCl (20 mM, pH 8·8), KCl (10 mM), (NH4)2SO4 (6 mM), MgSO4 (2 mM), Triton X-100 (0·1 %) and bovine serum albumin (0·1 mg ml–1). The template for PCR was either purified DNA (5–10 ng) or cells (~0·5 µl) on a 10 µl tip dipped into the PCR mix several times. Thermocycling consisted of initial denaturation (94 °C, 5 min), then 30 amplification cycles (94 °C for 30 s, 65 °C for 30 s, 72 °C for 1·5 min per kb) and a final extension (72 °C for 5 min).

Construction of pUS23 reporter plasmid.
The reporter plasmid pUS23 was constructed by PCR in E. coli JM109 using the pUC19 vector backbone. A fragment including the aadB 59-be was amplified from pMAQ105 with the primers NVC27/NVC28 and cloned into XbaI/PstI-cut pUC19 to yield pUS21. The aadB gene (NVC33/NVC34 amplicon from pMAQ105) and gfpmut3 gene (NVC35/NVC36 amplicon from pTGFP2) were digested with PstI/KpnI and KpnI/HindIII, respectively, and joined to PstI/HindIII-digested pUS21 in a three-way ligation, yielding pUS22. The rrnBT1T2 terminators from the JM109 chromosome were amplified with NVC37/NVC38, digested with XbaI/SalI and ligated into XbaI/SalI-cut pUS22 to yield pUS23 (Fig. 1). The expected structure and function of pUS23 in E. coli were confirmed by restriction mapping, sequencing of the reporter region and phenotypic tests (Gm MIC for aadB; fluorescence microscopy for gfp).

Introduction of pUS23 into P. stutzeri Q and selection of recombinants.
Plasmid pUS23 (1 µg) was electroporated into P. stutzeri Q (5x109 cells) and after recovery in 1 ml LB broth (2 h), dilutions were plated on LB-Ap. After 4 days, ApR colonies from one plate (several hundred) were pooled into 5 ml LB-Ap broth and grown to stationary phase (2 days, approx. 3x109 cells ml–1). Dilutions were spread on LB-Gm plates to detect activation of aadB and gfpmut3, and on LB plates to determine total viable count. Plain LB medium was used for determining the total count of P. stutzeri Q(pUS23) cells due to the instability of the intermediate P. stutzeri Q(pUS23) strain and its failure to reliably form single colonies on LB-Ap plates. Colonies appearing on Gm plates were subcultured to patches on the same medium and screened by PCR (below). The whole procedure from electroporation to GmR screening was repeated three times and the mean frequencies of transformation to ApR (per µg DNA) and conversion to GmR (per P. stutzeri Q cell) were calculated. Under the conditions used, the spontaneous frequencies of ApR and GmR in controls without plasmids were 3·6x10–7 per cell and <10–10 per cell, respectively (mean of three experiments).

PCR screening for attI recombination.
ApR GmR colonies were initially screened for the presence of the attIPstQ-aadB integron recombination junction by PCR using the primers NVC70/NVC71. This separated the transformants into two GmR groups, those that putatively contained pUS23 integrated at attIPstQ and those that did not. A representative GmR strain of each type from all three experiments was purified by restreaking, yielding six ApR GmR P. stutzeri Q(pUS23) derivatives (strains Q23-7, Q23-10, Q23-12, Q23-13, Q23-17 and Q23-25) for further analysis. A second PCR screen for the aadB 59be-BGC001 junction was performed on these six strains using the primers NVC27/NVC75. Where evidence of pUS23 integration was obtained, both junction sequences were determined by purification and direct sequencing of the PCR products.

Hybridization analysis of aadB copy number.
Genomic DNA samples (2 µg) from P. stutzeri Q derivatives were digested for 3 h with 20 U BssHII, EcoRI, NcoI, PstI, SalI, SmaI or StuI and run on 0·8 % agarose gels. Southern blotting was done by capillary transfer in 20x SSC buffer (Sambrook & Russell, 2001) onto positive nylon membranes (Hybond-N+; Amersham Biosciences) with subsequent hybridization at high stringency (68 °C) according to the DIG kit instructions (Roche). The probe consisted of the DIG-dUTP-labelled aadB gene (573 bp), which was PCR-amplified (primers NVC33/NVC34) from a gel-purified PstI–KpnI fragment of pUS23. The CDP-Star reagent (Roche Applied Science) was used for detection.

Cloning and recovery of integrated reporter plasmid.
Based on hybridization with aadB, a 12 kb StuI genomic fragment from strain Q23-17 was excised from agarose, purified, ligated to EcoRV-digested, alkaline-phosphatase-treated pACYC184 and electroporated into JM109. Resultant CmR colonies were screened for loss of TcR by patching to LB-Cm and LB-Tc plates, then TcS clones were screened by PCR using NVC70/NVC71 to detect attI-aadB junctions. One positive clone was retained and designated pUS40. Plasmids were extracted from a 200 ml LB-Cm-Ap culture of E. coli JM109(pUS40) and used for sequencing with pUS23 construction primers, pUC19 vector primers and P. stutzeri Q integron primers (Table 2).

Analysis of ApR plasmids in P. stutzeri Q derivatives.
Total genomic DNA was extracted from the ApR GmR P. stutzeri Q(pUS23) strains Q23-7, Q23-17 and Q23-25, and the parental ApR P. stutzeri Q(pUS23) strain. DNA from each strain (10 µg) was electroporated into JM109 and, after recovery in LB broth, the cells were plated on either Ap or Gm medium. The mean ApR and GmR transformation frequencies were calculated from three experiments. From each set of ApR JM109 transformants, five clones were subcultured in LB-Ap broth, and the plasmids extracted and analysed by restriction-mapping (EcoRI/HindIII digest). An insertion in one plasmid was further analysed by sequencing using primers NVC113/NVC114.

Detection of aadB expression by RT-PCR.
RNA was extracted from late-exponential-phase cultures of P. stutzeri strain Q23-17 using the RNeasy kit (Qiagen) according to the manufacturer's instructions, with the following modifications. Approximately 2x109 cells (2 ml culture at OD600=1·0) were pelleted from broth and resuspended directly in buffer RLT, omitting lysozyme treatment. DNase incubation (37 °C, 30 min) was done with the RNA eluate rather than on-column and was followed by a second round of column purification. cDNA was prepared with Omniscript RTase (Qiagen) using 1 µg RNA and primer NVC108. Subsequent PCR reactions used 1 µl RT mixture with primers NVC91/NVC71 and AJH21/NVC71, and were performed under the same conditions as other PCRs (above). Negative controls contained either no RT enzyme or DNA from wild-type P. stutzeri Q, while a positive control contained DNA from strain Q23-17.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Construction and characterization of reporter pUS23 in E. coli
The reporter plasmid pUS23 (Fig. 1) mimics a gene cassette in that it is a circular element and includes promoterless genes (aadB Gm resistance and gfpmut3 green fluorescence) immediately downstream of a 59-be recombination site (aadB 59-be). It is distinct from naturally occurring gene cassettes in that it contains a plasmid replication origin, a resistance gene (ApR) with its own promoter and transcription terminators (rrnBT1T2) upstream of the aadB 59-be. Addition of the terminators was found to be necessary to ensure silence of the GmR and green fluorescence genes in JM109 (data not shown). In E. coli, pUS23 is replicative and has the phenotype ApR GmS GFP (Table 1). The sequence of pUS23 has been deposited in GenBank (AY894753). The narrow host range of pUC vectors means pUS23 is expected to be non-replicative in most bacteria. In cells containing functional integrons, capture of pUS23 could potentially give Ap resistance, Gm resistance and green fluorescence. In the case of ApR, this phenotype could result from integron-mediated recombination alone (at attI, 59-be or a secondary site), while in the cases of GmR and green fluorescence, the phenotypes require both recombination (at attI) and expression (from Pc).

Activation of cassette-associated phenotypes in P. stutzeri cells
Three criteria are necessary for activation of gcORFs. These are uptake of a cassette by transformation, capture of the cassette by a replicon within the cell and transcriptional activation of the gcORFs. Electroporation of pUS23 into P. stutzeri Q resulted in an unexpectedly high number of colonies on Ap plates [mean 6·2x104 c.f.u. (µg plasmid)–1 over two experiments], but none on Gm plates. Since individual colonies from Ap plates were unable to be subcultured onto LB-Ap plates it is probable this does not reflect chromosomal integration of the plasmid (e.g. at attIPstQ). When DNA was extracted from the total pool of P. stutzeri Q(pUS23) transformant colonies, plasmid bands were not visible in agarose gels (data not shown), but electroporation of this DNA into E. coli JM109 did yield ApR JM109 transformants at low frequencies [approx. 20 c.f.u. (µg DNA)–1]. A possible explanation is that the founder cells for the ApR colonies were those in which pUS23 was initially present, but not effectively maintained during further growth of the colonies. We subsequently found one report of pUC vector maintenance in a P. stutzeri strain that provides support for this hypothesis (Pemberton & Penfold 1992).

Whilst high levels of initial transformation to ApR were observed, the lack of activation of pUS23 reporter phenotypes suggests that either the wild-type integron is not active under the laboratory growth conditions or the frequency of recombination was below the transformation frequencies attainable in our experiments. In previous reports on integron activity, the frequencies of cassette activation by wild-type class I integrons were low (10–7–10–9) (Hall et al., 1991; Recchia et al., 1994), but still well above the level of spontaneous GmR observed here (<10–10 per cell). We experimented with a number of variables aimed at increasing the likelihood of observing recombination events, including increasing the plasmid concentration and extending the time available for recombination to occur (Collis et al., 1993). A two-step approach, in which pooled ApR P. stutzeri Q(pUS23) transformants were subcultured in LB-Ap broth prior to plating out on to LB-Gm media was ultimately successful in yielding GmR colonies. In these experiments GmR colonies arose from the LB-Ap subculture at a frequency of 2·5x10–7 per c.f.u. (mean of three experiments).

At least three broad categories of event capable of activating aadB expression can be postulated; integron-related events where IntIPstQ mediates integration of pUS23 at attIPstQ and a Pc promoter directs expression of the genes; other chromosomal integration events where pUS23 is integrated downstream of a chromosomal promoter either by IntIPstQ or by other recombinases; and plasmid-related events whereby mutation or rearrangement of the plasmid leads to enhanced maintenance of the plasmid and activation of reporter genes.

Capture of pUS23 by InPstQ occurred by site-specific recombination
We expected that the majority of reporter gene activation events would be integron-related. Accordingly, PCR screening targeted the attI-aadB junction predicted to be formed as a result of integration of pUS23 at attIPstQ. Approximately 30 GmR colonies from each of three independent transformation experiments were screened and products of the expected size (425 bp) were found in 4/32, 2/23 and 1/27 cases. One PCR-positive strain from each experiment (Q23-7, Q23-17 and Q23-25; hereafter referred to as attI-GmR strains) was further tested by PCR-targeting the predicted aadB 59-be/BGC001 junction between InPstQ and pUS23, and a product of the expected size (509 bp) was observed (Fig. 2). The junction PCR products from the attI-GmR strains were sequenced and aligned to the known pUS23 and InPstQ sequences (Fig. 3). Crossovers located within the 1R core site (GTTAGGC) of the aadB 59-be (Stokes et al., 1997) and the comparable core site of attIPstQ (also GTTAGGC) were observed in all cases, indicating that the reporter plasmid had integrated via site-specific recombination, most likely mediated by IntIPstQ. Therefore, on average, 8 % of all GmR colonies contained pUS23 integrated at attIPstQ. Based on the frequency of appearance of all types of GmR colonies (2·5x10–7 per c.f.u.; see above), the overall frequency of attI integration can be calculated to be 2·0x10–8 per cell.



View larger version (91K):
[in this window]
[in a new window]
 
Fig. 2. PCR and fluorescence analysis of ApR GmR P. stutzeri Q(pUS23) derivatives. (a) PCR of attI-aadB junction (NVC70/NVC71); expected product is 425 bp. (b) Green fluorescence of cells under blue illumination. (c) PCR of aadB 59-be-BGC001 junction (NVC27/NVC75); expected product is 509 bp. Strain designations are listed across the top of the figure. Negative controls consisted of JM109(pUS23) and wild-type P. stutzeri Q.

 


View larger version (40K):
[in this window]
[in a new window]
 
Fig. 3. Amplified attI-pUS23 junctions in ApR GmR strains Q23-7, Q23-17 and Q23-25. (a) Junctions between attIPstQ and aadB. (b) Junctions between aadB 59-be and P. stutzeri Q BGC001. The junctions from strain Q23-17, sequenced from a subcloned chromosomal fragment (pUS40), were identical to those shown here. Bases conserved across all positions are shaded black, bases conserved at >75 % of positions are shaded grey and non-conserved positions have no shading.

 
The phenotype of the attI-GmR strains was further investigated. The MIC for Gm ranged from 15 to 20 µg ml–1, while the MIC for wild-type P. stutzeri Q was 3 µg ml–1. All attI-GmR strains also showed strong ApR (MIC >100 µg ml–1), and both resistances were stably maintained after at least one round of subculture on antibiotic-free medium, in contrast to the unstable ApR of the initial P. stutzeri Q(pUS23) transformants (data not shown). All the attI-GmR strains were green-fluorescent by microscopy (Fig. 2), confirming that both reporter genes had been activated and that GmR was unlikely to be due to a spontaneous chromosomal mutation. The stabilization of ApR, and the appearance of GmR and fluorescence are all consistent with integron capture and expression of the pUS23 gene cassette.

Detection of aadB copies by Southern hybridization
Since a large proportion of GmR strains did not involve integration of pUS23 at attIPstQ, we tested for the possibility of additional pUS23 integration events in the attI-GmR strains. Genomic DNA from the three attI-GmR strains was extracted, digested with PstI and Southern blots were probed with the aadB gene (Fig. 4). A unique PstI site is situated upstream of aadB in the reporter plasmid and a PstI site is found in the third P. stutzeri Q gene cassette (BGC003), and thus we expected to observe a band of 5·5 kb representing attI-integrated pUS23, and possibly also a band at 4·3 kb representing free plasmid. The hybridization pattern of strain Q23-7 was consistent with the presence of both free and attI-integrated pUS23 (Fig. 4). Strain Q23-25 was also predicted to contain free pUS23 (4·3 kb band), but an inferred attI-integrated band was at 6·6 kb rather than the expected 5·5 kb.



View larger version (47K):
[in this window]
[in a new window]
 
Fig. 4. Hybridization of aadB gene with DNA from GmR P. stutzeri Q derivatives. (a) PstI digests of total genomic DNA from three attI-GmR strains. Negative control is wild-type P. stutzeri Q DNA and positive control is E. coli JM109(pUS23) genomic DNA. (b) Various digests of Q23-17 DNA. Negative control is wild-type P. stutzeri Q DNA and positive control is PstI-digested E. coli JM109(pUS23) genomic DNA. Hybridizing fragments from EcoRI, NcoI, PstI, SmaI and StuI are approximately 1 kb larger than predicted from the sequences of InPstQ and pUS23, while BssHII and SalI fragments are smaller than predicted; the NcoI fragment matched the expected size.

 
In strain Q23-17 the only band detected was an inferred attI-integrated copy, but again this was at 6·6 kb, rather than the 5·5 kb predicted for the original pUS23. Only one copy of aadB was present in strain Q23-17 based on Southern hybridizations with DNA digested with a wide range of enzymes (Fig. 4b), but bands observed in the digests were in almost all cases of different sizes to those predicted from the pUS23 and InPstQ sequences. In the absence of digestion, a hybridization signal was seen with chromosomal DNA, consistent with chromosomal integration rather than free reporter plasmid. Taken together with the junction sequencing data presented above, the source of the GmR and green fluorescence phenotypes in strain Q23-17 appeared to be a derivative form of the pUS23 gene cassette with an approximately 1 kb insertion integrated at the attI site of InPstQ.

Sequence analysis of integrated pUS23 and flanking integron DNA from strain Q23-17
Due to the unexpected complexity of the Southern blot data and the implied presence of pUS23 and derivative plasmids in both free and integrated forms, questions relating to pUS23 structure, genomic location and reporter gene expression remained. To address these questions, we cloned and sequenced a 12 kb StuI fragment of strain Q23-17 DNA containing pUS23 with flanking chromosomal DNA on either side (pUS40). Extensive sequencing of the StuI fragment confirmed the integration of a derivative form of the pUS23 gene cassette by site-specific recombination at an attI site. Through PCR, partial sequence analysis and restriction mapping of pUS40, this site was confirmed to be in the previously described integron InPstQ (Holmes et al., 2003b). The flanking sequences of the StuI insert included the expected 167 bp of intIPstQ at one end and 586 bp of the P. stutzeri Q chromosomal gene orf136 at the other end of the insert DNA. PCR and restriction mapping indicated that the 10 gene cassettes of the wild-type InPstQ array were present in the StuI fragment in their original order (data not shown).

The near-complete sequence of the integrated pUS23 derivative in the cloned StuI fragment was obtained and the only sequence variation relative to the initial construct, pUS23, was the presence of an insertion sequence (IS) between bla and ori, inserted at TTAAAAGGAT^CTAGGTGAAG. The 1191 bp IS element in the integrated pUS23 was designated ISPst5 (GenBank AY894752) and is a member of the IS5 family. ISPst5 encodes a 326 aa DDE-type transposase, has 12 bp perfect terminal inverted repeats and creates a 4 bp target repeat (CTAG). ISPst5 has 99 % DNA identity to an IS associated with alkylbenzene degradation genes in Pseudomonas putida 01G3 (Chablain et al., 2001) and 83–86 % DNA identity to IS elements associated with naphthalene and carbazole catabolic plasmids in other Pseudomonas strains (Dennis & Zylstra, 2004; Nojiri et al., 2001). Since IS elements may have pleiotropic effects via promoters reading out from the IS (Galas & Chandler, 1989), we examined in more detail the expression of reporter genes in the attI-GmR strain Q23-17.

Expression of aadB and gfp is from a Pc equivalent promoter in InPstQ
The orientation of attI-integrated pUS23 in strain Q23-17 was as shown in Fig. 1(c) and provided good initial evidence that GmR and green fluorescence in this strain were due to integron-mediated activation of the reporter genes, independently of any effect of the IS. Previous studies have localized Pc in class 1 and class 3 integrons to within the 5' end of the intI gene (the –10 region corresponds to the Ser-30 codon of intI1). A Pc equivalent in P. stutzeri Q is not found at precisely the same site, but a putative {sigma}70-type promoter does exist nearby (the –10 region corresponds to Asp-18 of intI1). RT-PCR was used to test the functionality of the promoter in strain Q23-17, using a reverse primer (NVC71) within aadB and forward primers just upstream (AJH21) or just downstream (NVC91) of the promoter sequence (Fig. 5a), resulting in amplification of various intI-aadB fragments (Fig. 5b). A strong product of the expected size was seen with NVC91/NVC71 from reverse-transcribed RNA of Q23-17, but no product was obtained in the absence of RT. A similar pattern was observed with AJH21/NVC71, except that the product from RNA was very weak. The data confirmed that initiation of aadB transcription was occurring from within InPstQ and suggest that almost all of this transcription was due to the putative Pc promoter TTGAGC-n16-TCTGAT found within the 5' end of the intIPstQ gene (Fig. 5a). Essentially identical results were seen with RNA from strain Q23-17 cells grown in plain LB medium (data not shown), indicating that expression from Pc was not induced by antibiotics and is likely to be constitutive.



View larger version (60K):
[in this window]
[in a new window]
 
Fig. 5. Localization of Pc promoter by RT-PCR analysis of aadB gene expression in strain Q23-17. (a) Schematic diagram of attI region of InPstQ. Primer binding sites and restriction sites are underlined, while promoter elements (–35 and –10 regions), the intIPstQ start codon, the attIPstQ-aadB junction (GTTAGGC), the aadB ribosome-binding site (RBS) and start codon are shown in bold. The binding site of reverse primer NVC71 (not shown) is approximately 200 bp downstream of the aadB start codon. (b) RT-PCR data showing amplification with RNA and DNA templates from strain Q23-17 (shown as ‘17’) and P. stutzeri Q (shown as ‘Q’) using two alternative primer pairs. Reactions with and without reverse transcriptase are designated +RT and –RT, respectively.

 
Characterization of original and derivative pUS23 plasmids in GmR P. stutzeri Q strains
Based on hybridization data (Fig. 4), two of the three attI-GmR strains appeared to contain free plasmid. While strong evidence was obtained that integration of pUS23 at attI was responsible for the GmR phenotype of Q23-17, it was possible that plasmid rearrangements, mutations or insertion sequences contributed to GmR in the other two attI-GmR strains. To examine this possibility, genomic DNA from the attI-GmR strains was electroporated into E. coli JM109. These experiments yielded ApR colonies from Q23-7 [49±13 c.f.u. (µg DNA)–1] and Q23-25 [52±21 c.f.u. (µg DNA)–1], but not from Q23-17, consistent with the Southern hybridization data (Fig. 4). None of the experiments yielded GmR transformants of E. coli.

Restriction mapping of five recovered plasmid clones from Q23-7 and Q23-25 indicated an insertion of 1·2 kb in one plasmid from Q23-25, while all the other plasmids matched the profile of the original pUS23 (data not shown). Sequencing revealed that the insertion was ISPst5, in an identical location to that seen in the attI-integrated pUS23 copy in strain Q23-17 (between bla and ori in the reporter backbone). The fact that two independent isolates contained ISPst5 inserted at identical plasmid locations suggests a degree of target specificity for this insertion sequence, which is consistent with data available for IS5 family IS elements, which preferentially integrate at the sequence YTAR (usually CTAG, as observed with ISPst5) (Mahillon & Chandler, 1998). None of the E. coli JM109 transformants harbouring recovered pUS23, or the derivative pUS23-ISPst5, were GmR. Retransformation of the pUS23-ISPst5 plasmid from Q23-25 back to wild-type P. stutzeri Q confirmed that this plasmid conferred ApR, but not GmR, indicating that changes in the reporter plasmid itself were not responsible for GmR and that the activation of the cassette-associated genes occurred via the CI.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The significance of CIs
The gene cassette metagenome is a major source of genetic novelty. Through their manipulation of this resource, integrons are expected to have significant evolutionary impact. Based on the class 1 integron paradigm, where diverse antibiotic resistance genes were assembled in less than 50 years, natural selection for advantageous combinations of genes can occur relatively quickly. CIs are particularly interesting in this regard, since a particular integron backbone may operate within a bacterial lineage over evolutionarily significant periods of time (Rowe-Magnus et al., 2003; Gillings et al., 2005). It is plausible that CIs have been continually acquiring gene cassettes, expressing their associated genes and rearranging their transcriptional order throughout the evolution of certain genera. In the cases of Vibrio and Xanthomonas, CIs are inferred to have been resident over thousands of years, ample time for selection of beneficial gene combinations. But do CIs have the same capabilities as mobilized integrons? MIs show high diversity in 59-be sites and their associated gcORFs have a tendency to exhibit a functional theme. It seems reasonable to infer that the apparent linkage between gcORF function and 59-be diversity reflects natural selection for gene cassettes from different sources that deliver related fitness advantages. In contrast, CIs share similarity of 59-be sites within a species and it seems unlikely that cassettes in CI arrays, particularly very large ones, share a functional theme. The fixed chromosomal location of CIs is expected to limit their access to diverse sources of gene cassettes, but do these patterns in CI and MI cassette arrays also reflect different abilities to facilitate natural selection for useful combinations of gene cassettes?

Two possibilities worthy of investigation are that not all integrons may have the full set of functions observed in resistance integrons, or that all integrons have the same functions but differ in their level of activity. For example limitation or absence of gcORF expression would largely remove the selective mechanisms capable of leading to accumulation of gene cassettes from different backgrounds (containing diverse 59-be) or with functionally related phenotypes. Alternatively, higher recombination specificity and/or lower frequencies of recombination activity could also explain why CIs are not associated with readily detectable phenotypes since the chance of acquiring a selectively advantageous gene would be reduced. In summary two key questions for CIs are: do they express gcORFs, and is the rate and/or specificity of gene cassette capture limited with respect to the mobilized integrons? Answering these questions requires better sampling of integron diversity and methods to look at integron activity in a natural context. Here we have used a reporter gene cassette to address these questions.

The InPstQ CI can activate gene cassettes
Our data are the first demonstration that native-form CIs can capture gene cassettes and express cassette-associated genes, and greatly extend previous observations that CIs can act as accessible reservoirs of gene cassettes (Rowe-Magnus et al., 2002). Cloning, sequencing, hybridization and RT-PCR data confirmed that integron activity resulted in gcORF expression. Hybridization and sequence data showed that in three independent strains pUS23 was captured by site-specific recombination between attIPstQ and the aadB 59-be of pUS23. In strain Q23-17 subsequent activation of the aadB and gfp genes was shown to be due to expression from a promoter located within InPstQ. We conclude that InPstQ is a fully functional integron whose activities are sufficient to enable the capture of gene cassettes and expression of associated genes. While we cannot extrapolate from our study of InPstQ to infer functions of other CIs, it should be possible to use pUS23 or similar reporter gene cassettes to rapidly screen other CIs for function.

Interpretation of the data was complicated by the presence of an insertion sequence in the reporter cassette. Examples of both original and IS-derivative pUS23 were rescued by transformation into E. coli JM109. The derivative form was shown to have a GmS GFP phenotype in E. coli and when transformed back into wild-type P. stutzeri. These data show that ISPst5 does not cause the GmR GFP+ phenotype. The association between ISPst5 and pUS23 is also highly unlikely to be a prerequisite for capture of the pUS23 cassette by InPstQ, since one strain (Q23-7) appeared to contain only the original pUS23. However, the interaction between ISPst5 and pUS23 may have had other effects (see below).

Interaction between pUS23 and ISPst5 is likely to have influenced the frequency of cassette capture
The other outstanding issues with respect to CI function are the rate and specificity of gene cassette capture. With respect to 59-be specificity, it is notable that InPstQ was able to recognize the aadB 59-be, a recombination site that is divergent from the 59-be family typically observed in Pseudomonas CI arrays (Holmes et al., 2003b; Vaisvila et al., 2001). We did not investigate this further here. Our estimate (2x10–8 per cell) of the frequency of cassette capture and activation by InPstQ is of a similar order of magnitude to frequencies of cassette phenotype activation observed with wild-type class I integrons (Hall et al., 1991; Recchia et al., 1994). However, these numbers are not directly comparable due to different methodologies. In addition, the effect of experimental conditions such as antibiotic exposure is unknown (Beaber et al., 2004) and the data are further complicated by the fact that pUS23 did not behave as a non-replicative gene cassette in our system.

P. stutzeri is an exceedingly diverse species complex (Cladera et al., 2004) and there are reports of ColE1-like replicons being replicative in some strains and non-replicative in others (Pemberton & Penfold, 1992). Here, pUS23 was found to be maintained to some extent in P. stutzeri Q. It is probable that this increased the chance of recombination between the aadB 59-be and attIPstQ, thus increasing the frequency of observation of cassette capture. Surprisingly, this effect was apparently exacerbated by ISPst5 insertion. The finding that two independently isolated pUS23 plasmid derivatives contained ISPst5 inserted upstream of ori in the pUC19 backbone raises the possibility that the insertion may have stimulated the origin of replication, which was initially only weakly functional in P. stutzeri Q. Activation of the pBR322 plasmid origin by an IS-derived promoter has been observed with Tn5 (Lupski et al., 1986). In any event the maintenance of the cassette as a plasmid allowed us to detect low frequency events that may otherwise have not been observed with a non-replicative cassette.

Are there alternate routes for gene cassette activation?
It is widely thought that integrons are the major route by which cells access the gene cassette metagenome. A surprising outcome of our study was that less than 10 % of GmR P. stutzeri Q colonies resulted from integration at attIPstQ (we term the remainder unknown-GmR strains). Southern hybridization revealed that pUS23 was at heterogeneous genomic locations in three representative unknown-GmR strains (data not shown), suggesting that multiple routes for pUS23 reporter activation exist. These could include alternate integron sites or recombination involving insertion sequences.

It is unlikely that pUS23 initially contained sequences that could support homologous recombination with the P. stutzeri chromosome, and Southern blots (Fig. 4) confirmed that aadB was absent from wild-type P. stutzeri Q. However, insertion of ISPst5 into pUS23 would provide a convenient target for homologous recombination with other chromosomally borne copies of the same IS. Similar reactions are believed to be responsible for the integration of the F plasmid into the E. coli chromosome and subsequent production of Hfr strains (Galas & Chandler, 1989). In the case of the unknown-GmR strains such events would also need to have resulted in expression of aadB and gfp.

Integrons show a strong preference for capture of gene cassettes by recombination at their cognate attI site (Collis et al., 2002b), although cassette capture by recombination at 59-be sites or secondary sites (not associated with integrons or gene cassettes) is known (Francia et al., 1993; Recchia et al., 1994) and expression from promoters upstream of secondary sites has been demonstrated (Segal & Elisha, 1999). We screened three of the unknown-GmR strains (Q23-10, Q23-12 and Q23-13) for insertion of pUS23 at the 59-be sites associated with the 10 gene cassettes in the InPstQ array by PCR (data not shown), but the results indicated that pUS23 had not integrated into the known cassette array. Another possibility is that attI, 59-be or secondary sites may exist elsewhere in the chromosome. For example, Pseudomonas alcaligenes harbours multiple cassette arrays (Vaisvila et al., 2001), and the genomes of Shewanella (GenBank NC_004347) and Treponema (GenBank NC_002967) contain gene cassettes at locations isolated from the known integron. IntI-mediated integration of gene cassettes at secondary sites has been reported. If alternate attI or 59-be sites were the site of integration, then we must also account for the low proportion of capture at the attI site of InPstQ, as previous studies have suggested the cognate attI site is strongly preferred over 59-be sites (Collis et al., 2002a, b).

The location of Pc is worth noting in the context of integration site preference. The implied overlap of intI and cassette transcripts suggests the potential for a regulatory relationship between expression of these two integron elements. This makes ‘biological sense’ in that if a cell is expressing an advantageous combination of gcORFs, concomitant IntI activity would potentially result in disruption. Previous studies on recombination site preferences have all involved assays where the integrase was supplied in trans and not under the control of its native promoter, while in our system the native integron was assayed for function. If there is a regulatory relationship between intI expression and Pc-directed gcORF expression then this could influence the observed outcomes. This is worthy of further investigation.

The pUS23 reporter as a tool for integron recovery
Since the gene cassette metagenome is tightly linked to the activity of integrons, understanding integron distribution and diversity is a key element in determining the evolutionary significance of gene cassettes. In this context, another application for the pUS23 reporter cassette is in the recovery of new integrons. Screening for resistance phenotypes clearly does not recover representative integron diversity (Barlow et al., 2004) and PCR-based screening strategies (Nemergut et al., 2004; Nield et al., 2001) are limited to recovery of integron fragments. Useful features for second-generation reporter constructs may include broad host-range (Kovach et al., 1995) or conditional (Herrero et al., 1990) replication origins, and mobilization functions compatible with promiscuous E. coli delivery vehicles (Simon et al., 1983). A reporter-based screening strategy would have the advantage of detecting functional integrons and allow recovery of the intact integron in its native host. Application of this approach to environmental samples could yield valuable information on the in situ activity and ecological relevance of various integron types.


   ACKNOWLEDGEMENTS
 
We thank Hatch Stokes, Lars Hansen, Gordon Stevenson and Harold Erickson for generously providing plasmids and strains. Ruth Hall and Hatch Stokes are acknowledged for helpful comments on the manuscript. N. V. C. was supported by a University of Sydney Sesqui postdoctoral fellowship. Research in the A. J. H. lab was supported by grants from the Australian Research Council.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Barlow, R. S., Pemberton, J. M., Desmarchelier, P. M. & Gobius, K. S. (2004). Isolation and characterization of integron-containing bacteria without antibiotic selection. Antimicrob Agents Chemother 48, 838–842.[Abstract/Free Full Text]

Beaber, J. W., Hochhut, B. & Waldor, M. K. (2004). SOS response promotes horizontal dissemination of antibiotic resistance genes. Nature 427, 72–74.[CrossRef][Medline]

Boyd, D., Cloeckaert, A., Chaslus-Dancla, E. & Mulvey, M. R. (2002). Characterization of variant Salmonella genomic island 1 multidrug resistance regions from serovars Typhimurium DT104 and Agona. Antimicrob Agents Chemother 46, 1714–1722.[Abstract/Free Full Text]

Bunny, K. L., Hall, R. M. & Stokes, H. W. (1995). New mobile gene cassettes containing an aminoglycoside resistance gene, aacA7, and a chloramphenicol resistance gene, catB3, in an integron in pBWH301. Antimicrob Agents Chemother 39, 686–693.[Abstract]

Chablain, P. A., Zgoda, A. L., Sarde, C. O. & Truffaut, N. (2001). Genetic and molecular organization of the alkylbenzene catabolism operon in the psychrotrophic strain Pseudomonas putida 01G3. Appl Environ Microbiol 67, 453–458.[Abstract/Free Full Text]

Chang, A. C. Y. & Cohen, S. N. (1978). Construction and characterization of amplifiable multicopy DNA cloning vehicles derived from the P15A cryptic miniplasmid. J Bacteriol 134, 1141–1156.[Medline]

Cladera, A. M., Bennasar, A., Barcelo, M., Lalucat, J. & Garcia-Valdes, E. (2004). Comparative genetic diversity of Pseudomonas stutzeri genomovars, clonal structure, and phylogeny of the species. J Bacteriol 186, 5239–5248.[Abstract/Free Full Text]

Collis, C. M. & Hall, R. M. (1995). Expression of antibiotic resistance genes in the integrated cassettes of integrons. Antimicrob Agents Chemother 39, 155–162.[Abstract]

Collis, C. M., Grammaticopoulos, G., Briton, J., Stokes, H. W. & Hall, R. M. (1993). Site-specific insertion of gene cassettes into integrons. Mol Microbiol 9, 41–52.[Medline]

Collis, C. M., Recchia, G. D., Kim, M. J., Stokes, H. W. & Hall, R. M. (2001). Efficiency of recombination reactions catalyzed by class 1 integron integrase IntI1. J Bacteriol 183, 2535–2542.[Abstract/Free Full Text]

Collis, C. M., Kim, M. J., Partridge, S. R., Stokes, H. W. & Hall, R. M. (2002a). Characterization of the class 3 integron and the site-specific recombination system it determines. J Bacteriol 184, 3017–3026.[Abstract/Free Full Text]

Collis, C. M., Kim, M. J., Stokes, H. W. & Hall, R. M. (2002b). Integron-encoded IntI integrases preferentially recognize the adjacent cognate attI site in recombination with a 59-be site. Mol Microbiol 46, 1415–1427.[CrossRef][Medline]

Dennis, J. J. & Zylstra, G. J. (2004). Complete sequence and genetic organization of pDTG1, the 83 kilobase naphthalene degradation plasmid from Pseudomonas putida strain NCIB 9816-4. J Mol Biol 341, 753–768.[CrossRef][Medline]

Drouin, F., Melancon, J. & Roy, P. H. (2002). The intI-like tyrosine recombinase of Shewanella oneidensis is active as an integron integrase. J Bacteriol 184, 1811–1815.[Abstract/Free Full Text]

Francia, M. V., Delacruz, F. & Lobo, J. M. G. (1993). Secondary sites for integration mediated by the Tn21 integrase. Mol Microbiol 10, 823–828.[Medline]

Galas, D. J. & Chandler, M. (1989). Bacterial insertion sequences. In Mobile DNA, pp. 109–162. Edited by D. Berg & M. Howe. Washington DC: American Society for Microbiology.

Gillings, M. R., Holley, M. P., Stokes, H. W. & Holmes, A. J. (2005). Integrons in Xanthomonas: a source of species genome diversity. Proc Natl Acad Sci U S A 102, 4419–4424.[Abstract/Free Full Text]

Hall, R. M. & Collis, C. M. (1995). Mobile gene cassettes and integrons: capture and spread of genes by site-specific recombination. Mol Microbiol 15, 593–600.[CrossRef][Medline]

Hall, R. M. & Collis, C. M. (1998). Antibiotic resistance in gram-negative bacteria: the role of gene cassettes and integrons. Drug Resist Updat 1, 109–119.

Hall, R. M., Brookes, D. E. & Stokes, H. W. (1991). Site-specific insertion of genes into integrons – role of the 59-base element and determination of the recombination cross-over point. Mol Microbiol 5, 1941–1959.[Medline]

Hanau-Bercot, B., Podglajen, I., Casin, I. & Collatz, E. (2002). An intrinsic control element for translational initiation in class 1 integrons. Mol Microbiol 44, 119–130.[CrossRef][Medline]

Hansen, L. H., Ferrari, B., Sorensen, A. H., Veal, D. & Sorensen, S. J. (2001). Detection of oxytetracycline production by Streptomyces rimosus in soil microcosms by combining whole-cell biosensors and flow cytometry. Appl Environ Microbiol 67, 239–244.[Abstract/Free Full Text]

Herrero, M., de Lorenzo, V. & Timmis, K. N. (1990). Transposon vectors containing non-antibiotic resistance selection markers for cloning and stable chromosomal insertion of foreign genes in gram-negative bacteria. J Bacteriol 172, 6557–6567.[Medline]

Hochhut, B., Lotfi, Y., Mazel, D., Faruque, S. M., Woodgate, R. & Waldor, M. K. (2001). Molecular analysis of antibiotic resistance gene clusters in Vibrio cholerae O139 and O1 SXT constins. Antimicrob Agents Chemother 45, 2991–3000.[Abstract/Free Full Text]

Holmes, A. J., Gillings, M. R., Nield, B. S., Mabbutt, B. C., Nevalainen, K. M. H. & Stokes, H. W. (2003a). The gene cassette metagenome is a basic resource for bacterial genome evolution. Environ Microbiol 5, 383–394.[CrossRef][Medline]

Holmes, A. J., Holley, M. P., Mahon, A., Nield, B., Gillings, M. & Stokes, H. W. (2003b). Recombination activity of a distinctive integron-gene cassette system associated with Pseudomonas stutzeri populations in soil. J Bacteriol 185, 918–928.[Abstract/Free Full Text]

Kovach, M. E., Elzer, P. H., Hill, D. S., Robertson, G. T., Farris, M. A., Roop, R. M. & Peterson, K. M. (1995). Four new derivatives of the broad-host-range cloning vector pBBRMCS, carrying different antibiotic-resistance cassettes. Gene 166, 175–176.[CrossRef][Medline]

Leon, G. & Roy, P. H. (2003). Excision and integration of cassettes by an integron integrase of Nitrosomonas europaea. J Bacteriol 185, 2036–2041.[Abstract/Free Full Text]

Levesque, C., Brassard, S., Lapointe, J. & Roy, P. H. (1994). Diversity and relative strength of tandem promoters for the antibiotic-resistance genes of several integrons. Gene 142, 49–54.[CrossRef][Medline]

Liebert, C. A., Hall, R. M. & Summers, A. O. (1999). Transposon Tn21, flagship of the floating genome. Microbiol Mol Biol Rev 63, 507–522.[Abstract/Free Full Text]

Lupski, J. R., Projan, S. J., Ozaki, L. S. & Godson, G. N. (1986). A temperature-dependent pBR322 copy number mutant resulting from a Tn5 position effect. Proc Natl Acad Sci U S A 83, 7381–7385.[Abstract/Free Full Text]

Mahillon, J. & Chandler, M. (1998). Insertion sequences. Microbiol Mol Biol Rev 62, 725–774.[Abstract/Free Full Text]

Michael, C. A., Gillings, M. R., Holmes, A. J., Hughes, L., Andrew, N. R., Holley, M. P. & Stokes, H. W. (2004). Mobile gene cassettes: a fundamental resource for bacterial evolution. Am Nat 164, 1–12.[CrossRef][Medline]

Nemergut, D. R., Martin, A. P. & Schmidt, S. K. (2004). Integron diversity in heavy-metal-contaminated mine tailings and inferences about integron evolution. Appl Environ Microbiol 70, 1160–1168.[Abstract/Free Full Text]

Nield, B. S., Holmes, A. J., Gillings, M. R., Recchia, G. D., Mabbutt, B. C., Nevalainen, K. M. & Stokes, H. W. (2001). Recovery of new integron classes from environmental DNA. FEMS Microbiol Lett 195, 59–65.[CrossRef][Medline]

Nojiri, H., Sekiguchi, H., Maeda, K., Urata, M., Nakai, S. I., Yoshida, T., Habe, H. & Omori, T. (2001). Genetic characterization and evolutionary implications of a car gene cluster in the carbazole degrader Pseudomonas sp. strain CA10. J Bacteriol 183, 3663–3679.[Abstract/Free Full Text]

Pemberton, J. M. & Penfold, R. J. (1992). High-frequency electroporation and maintenance of pUC-based and pBR-based cloning vectors in Pseudomonas stutzeri. Curr Microbiol 25, 25–29.[CrossRef][Medline]

Recchia, G. D., Stokes, H. W. & Hall, R. M. (1994). Characterization of specific and secondary recombination sites recognized by the integron DNA integrase. Nucleic Acids Res 22, 2071–2078.[Abstract]

Rowe-Magnus, D. A. & Mazel, D. (2001). Integrons: natural tools for bacterial genome evolution. Curr Opin Microbiol 4, 565–569.[CrossRef][Medline]

Rowe-Magnus, D. A., Guerout, A. M., Ploncard, P., Dychinco, B., Davies, J. & Mazel, D. (2001). The evolutionary history of chromosomal super-integrons provides an ancestry for multiresistant integrons. Proc Natl Acad Sci U S A 98, 652–657.[Abstract/Free Full Text]

Rowe-Magnus, D. A., Guerout, A. M. & Mazel, D. (2002). Bacterial resistance evolution by recruitment of super-integron gene cassettes. Mol Microbiol 43, 1657–1669.[CrossRef][Medline]

Rowe-Magnus, D. A., Guerout, A. M., Biskri, L., Bouige, P. & Mazel, D. (2003). Comparative analysis of superintegrons: engineering extensive genetic diversity in the Vibrionaceae. Genome Res 13, 428–442.[Abstract/Free Full Text]

Sambrook, J. & Russell, D. W. (2001). Molecular Cloning: a Laboratory Manual, 3rd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Segal, H. & Elisha, B. G. (1999). Characterization of the Acinetobacter plasmid, pRAY, and the identification of regulatory sequences upstream of an aadB gene cassette on this plasmid. Plasmid 42, 60–66.[CrossRef][Medline]

Simon, R., Priefer, U. & Puhler, A. (1983). A broad hostrange mobilization system for in vivo genetic-engineering – transposon mutagenesis in gram-negative bacteria. BioTechnology 1, 784–791.[CrossRef]

Stokes, H. W., O'Gorman, D. B., Recchia, G. D., Parsekhian, M. & Hall, R. M. (1997). Structure and function of 59-base element recombination sites associated with mobile gene cassettes. Mol Microbiol 26, 731–745.[CrossRef][Medline]

Stokes, H. W., Holmes, A. J., Nield, B. S., Holley, M. P., Nevalainen, K. M., Mabbutt, B. C. & Gillings, M. R. (2001). Gene cassette PCR: sequence-independent recovery of entire genes from environmental DNA. Appl Environ Microbiol 67, 5240–5246.[Abstract/Free Full Text]

Szczepanowski, R., Krahn, I., Linke, B., Goesmann, A., Puhler, A. & Schluter, A. (2004). Antibiotic multiresistance plasmid pRSB101 isolated from a wastewater treatment plant is related to plasmids residing in phytopathogenic bacteria and carries eight different resistance determinants including a multidrug transport system. Microbiology 150, 3613–3630.[CrossRef][Medline]

Vaisvila, R., Morgan, R. D., Posfai, J. & Raleigh, E. A. (2001). Discovery and distribution of super-integrons among pseudomonads. Mol Microbiol 42, 587–601.[CrossRef][Medline]

Yanisch-Perron, C., Vieira, J. & Messing, J. (1985). Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33, 103–119.[CrossRef][Medline]

Received 24 March 2005; accepted 24 March 2005.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Coleman, N. V.
Articles by Holmes, A. J.
Articles citing this Article
PubMed
PubMed Citation
Articles by Coleman, N. V.
Articles by Holmes, A. J.
Agricola
Articles by Coleman, N. V.
Articles by Holmes, A. J.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2005 Society for General Microbiology.