Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, May 2005, p. 1802-1807, Vol. 49, No. 5
0066-4804/05/$08.00+0 doi:10.1128/AAC.49.5.1802-1807.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Department of Biological Sciences, University of Warwick, Coventry CV4 7AL, United Kingdom,1 University of Birmingham, Division of Immunity and Infection, Edgbaston, Birmingham B15 2TT, United Kingdom2
Received 21 July 2004/ Returned for modification 12 September 2004/ Accepted 30 January 2005
|
|
|---|
1 (attenuated qacE) genes. QAC resistance was higher in isolates from reed bed samples, and class 1 integron incidence was significantly higher for populations that were preexposed to QACs. This is the first study to demonstrate that QAC selection in the natural environment has the potential to coselect for antibiotic resistance, as class 1 integrons are well-established vectors for cassette genes encoding antibiotic resistance. |
|
|---|
QAC resistance genes fall into two families. The qacA/B genes belong to the major facilitator superfamily and are only found in staphylococci on multiresistance plasmids (30). Other QAC resistance genes belong to the small multidrug resistance family and include qacC/D, now known as smr, qacE, qacE
1, qacF, qacG, qacH, and qacJ (3, 7, 15-17, 21, 23, 31, 32). qacE, qacE
1, qacF, and qacG have been identified on integrons, and the remaining genes have been identified on multiresistance plasmids in staphylococci (qacG is unusual in that it has been reported to be carried by staphylococcal multiresistance plasmids and class 1 integrons in gram-negative bacteria).
Class 1 integrons consist of a 5' conserved region consisting of an integrase gene encoding a site-specific recombinase (42), an attI site (28) where cassettes are integrated, and a promoter, Pant, that regulates the expression of gene cassettes (10). Gene cassettes contain a protein coding region and a recombination site known as a 59-be site which is responsible for the orientation of integration (9). The variable regions of class 1 integrons contain the cassette genes, to the right of which lies the 3' conserved region, which may have one of three different backbone structures (27). The first backbone type consists of a Tn402 (In16)-like arrangement consisting of a tni module containing three transposition genes and a resolvase gene. The second, In5 type consists of qacE
1, sul1, orf5, orf6, and a partial tni module, tni
, consisting of two transposition genes. The third, In4 type carries just qacE
1, sul1, orf5, and orf6. Integrons carrying the complete tni module are able to undergo self-transposition, and it is thought that the In5 and In4 types may also be able to move if the tni gene products are supplied in trans (27).
The role of class 1 integrons in conferring antibiotic resistance to clinical isolates of many bacterial strains is well documented (5, 14, 22, 38, 44). Studies of the incidence of class 1 integrons in bacterial pathogens associated with agriculture and fish farming, such as Escherichia coli and Aeromonas salmonicida, have also shown a link between integrons and antibiotic resistance (2, 41). Studies of the incidence of integrons in environmental bacteria have been undertaken by Rosser and Young (34), who studied the incidence of class 1 integrons in isolates from the Tay estuary, where 3.6% of isolates carried class 1 integrons. Nield and coworkers (25) identified three new integron classes in total community DNAs extracted from Australian soils by using primers for a conserved region of the integrase gene and the 59-be site. A similar approach using degenerate primers targeting the conserved regions of 59-be sites identified 164 gene cassettes, with the majority showing no relationship to known sequences.
Antibiotic and QAC resistance genes, e.g., qacE, are both carried on class 1 integrons, so selection for QAC resistance may cause coselection for antibiotic resistance. This provides a potential reservoir of antibiotic-resistant bacteria in QAC-polluted environments. In staphylococci, the qacA/B genes carried on multiresistance plasmids confer a low-level resistance to chlorhexidine and QACs, and it has been suggested that the introduction of chlorhexidine into clinical environments has resulted in the selection of staphylococci containing qacA carried on multiresistance plasmids (35).
For this study, resistance to ditallowdimethylammomium chloride (DTDMAC) and cetyltrimethylammonium bromide (CTAB) was assessed in environmental bacteria. Bacteria were isolated from a reed bed system (Fig. 1) used to remediate effluent from a textile mill with high QAC usage (wool finishing effluent only) and also from control sites with low QAC exposure. The incidence of class 1 integrons was also investigated to test the hypothesis that QAC exposure coselects for antibiotic resistance by selecting for class 1 integrons.
![]() View larger version (16K): [in a new window] |
FIG. 1. Diagram of mill and reed bed system. Rbi, reed bed inlet; RB1 to RB3, reed beds 1 to 3; Rbo, reed bed outlet.
|
|
|
|---|
Isolation. Viable counts were carried out in triplicate on nutrient agar (NA) containing 5 µg/ml or 50 µg/ml CTAB or DTDMAC or with no selection; all plates contained 50 µg/ml cycloheximide to reduce fungal growth. Plates were incubated at 26°C. Forty to 50 colonies encompassing the range of colony morphologies observed were picked for each sample and streaked onto NA plates.
PCR.
Isolates were grown overnight in nutrient broth at 26°C, and phenol-chloroform extraction was performed. DNA concentrations were determined by measuring the absorbance of the sample at 260 nm on a spectrophotometer and were standardized to 100 ng/µl. intA- and intB-specific primers (34) were used to amplify the class 1 integrase gene, and specific primers were used to amplify the qacE and qacE
1 genes (19). Integron-positive isolates and approximately 20 additional isolates from both contaminated and control sites were identified by PCR amplification and sequencing of 16S rRNAs by the use of pA-pH primers (11). Standard PCR conditions were used, as described by Rosser and Young (34) and Kazama et al. (19). Confirmation of the presence of qacE on a class 1 integron was achieved by using CASS1/CASS2 primers (34), which amplify the variable region of integrons priming the intI1 and qacE or qacE
1 genes. PCR products were detected by electrophoresis in a 1.5% agarose gel, sequenced by the use of Big Dye Terminator, version 3.1, chemistry (Applied Biosystems), and run on a 3100 genetic analyzer. To aid in the identification of some isolates, we used an API 20NE test kit (bioMerieux) according to the manufacturer's instructions, incubating the bacteria at 30°C. The identification of strains was carried out according to the instructions in the API 20NE identification manual.
Statistical analysis. Plate counts were expressed as resistance quotients (RQs), for which the counts with selection were expressed as percentages of the counts with no selection. Differences in class 1 integrase gene frequencies between the polluted and control isolates were tested for significance by use of a chi-square test for comparisons of two proportions (from independent samples).
|
|
|---|
![]() View larger version (42K): [in a new window] |
FIG. 2. Relative resistance to DTDMAC and CTAB at 5 µg/ml. (Top) Viable plate counts, with error bars giving standard deviations. (Bottom) RQs showing percentages of resistant bacteria in each sample. Rbi, reed bed inlet; RB1 to RB3, reed beds 1 to 3; r, root-associated sample material; Rbo, reed bed outlet; RSM, river sediment downstream of outfall; CW1 and CW2, Cotswold farmland soil; DW, Droitwich farmland soil; CR1 to CR3, Warwickshire farmland soil.
|
![]() View larger version (37K): [in a new window] |
FIG. 3. Relative resistance to DTDMAC and CTAB at 50 µg/ml. (Top) Viable plate counts, with error bars giving standard deviations. (Bottom) RQs showing percentages of resistant bacteria in each sample. Rbi, reed bed inlet; RB1 to RB3, reed beds 1 to 3; r, root-associated sample material; Rbo, reed bed outlet; RSM, river sediment downstream of outfall; CW1 and CW2, Cotswold farmland soil; DW, Droitwich farmland soil; CR1 to CR3, Warwickshire farmland soil (Cryfield).
|
PCR screening. The results of screening for intI1, the integrase gene carried on class 1 integrons, are shown in Table 1. For isolates from contaminated samples (including river sediment), 7.98% were integron positive, in contrast to 0% of control samples, which was a statistically significant difference, with a chi-square value of 19.604 and a P value of <0.0001 when a comparison of proportions test was applied. When restricted to the reed bed samples, the incidence increased to 14.9%, which was also significantly different from that of the control samples.
|
View this table: [in a new window] |
TABLE 1. Incidence of intI1, qacE, and qacE 1
|
1 genes, with 18 isolates testing positive for qacE and 0 testing positive for qacE
1 (Table 1). Putative qacE PCR products were sequenced and illustrated 100% similarity to known qacE genes. Amplification of the integron variable region by use of the CASS1/CASS2 primers produced an 1,800-bp product, indicating that 18/19 integrons carried the qacE gene. 16S rRNA PCRs and subsequent sequencing of a 700-bp fragment revealed that the 18 isolates carrying qacE showed the highest similarities to Pseudomonas sp. (3 isolates), Serratia sp. (2 isolates), and Aeromonas hydrophila (13 isolates) and that the isolate lacking qacE showed the highest similarity to Enterobacteriaceae. A full list of the species identified is given in Table 2. In many cases, the 16S rRNA sequences showed high similarities to those of Aeromonas hydrophila, Aeromonas salmonicida, and Vibrio parahaemolyticus. The taxonomy of the genus Aeromonas is not straightforward, with the so-called "A. hydrophila complex" also including the closely related A. salmonicida and Aeromonas bestiarum (40). A. hydrophila is motile, whereas A. salmonicida strains are generally nonmotile. The API 20NE test kit gave an identification of A. hydrophila with a 99.3% certainty. Microscopic observation revealed the cells to be highly motile rods, strengthening the identification of A. hydrophila. |
View this table: [in a new window] |
TABLE 2. Identification of isolates
|
|
|
|---|
Resistance to CTAB at both concentrations showed a clear linkage with the reed bed habitat. Isolates from the mill effluent flowing into the reed bed and from the effluent discharged from the reed bed illustrated much less resistance than other isolates, for the former because there was little time for selection to occur upstream of the reed beds and for the latter because resistant bacteria may be associated with root systems or biofilms. Bacteria are known to be less susceptible to QACs when growing in biofilms (6), and the low levels of resistant bacteria in the reed bed outlet effluent confirmed that resistant strains reside within the reed bed sediment. It is also possible that low counts in the reed bed outlet effluent can be accounted for by imperfect operation of the system, i.e., some effluent flowed over the surface of the reed bed, diluting that passing through the beds. Very little resistance was seen for control samples at 5 µg/ml CTAB, and almost none was seen at 50 µg/ml. Again, the integron frequency was generally higher for samples with populations with high RQs. Integron-positive isolates were screened for qacE and qacE
1, and 18/19 isolates from the reed bed samples carried qacE, a fully functional multidrug efflux gene carried on a class 1 integron that has been shown to confer increased resistance to a variety of QACs, including CTAB (31). The integron frequency was also highest in RB1, which had a higher effluent flow rate than RB2 and RB3. Rosser and Young (34) reported a 3.6% incidence of class 1 integrons in environmental coliforms, i.e., Pseudomonas spp. and Vibrio spp. collected from the River Tay estuary water, which was similar to the overall incidence of 3.8% seen for this study. However, it is clear that class 1 integrons were not evenly distributed among samples but were only found in reed bed samples exposed to QACs (14.9% of bacterial isolates from reed bed samples) and not in isolates from agricultural soil. The high incidence of qacE observed (approximately 95% of integrons) is higher than those observed in other studies of environmental bacteria, which showed incidences of 46% (34) and 0% (19). qacE confers resistance to QACs, and its prevalence probably represents selection by QACs contained in the mill effluent. Integrons from clinical isolates are characterized by possessing a 3' conserved region, usually containing sul1 and qacE
1, but this arrangement seems far less usual in environmental isolates, possibly due to less exposure to sulfonamides, synthetic antimicrobials that have historically been used extensively in medicine (34).
In addition to acquired resistance conferred by multidrug efflux genes such as qacE, selection for intrinsically resistant strains and innate resistance mechanisms is also likely to occur. Bacteria such as Acinetobacter calcoaceticus (Acinetobacter sp. isolates were isolated from reed bed samples) are known to produce emulsan, a polyanionic heteropolysaccharide bioemulsifier that provides the cell with a protective barrier against CTAB (39). Rhodococcus erythropolis, which was isolated from reed bed inlet and outlet samples, is also highly resistant to disinfectants (4). E. coli mutants produced spontaneously by repeated exposure to CTAB have been reported to develop multidrug resistance (18), and it was suggested that the mechanism of resistance is a combination of an altered lipopolysaccharide profile and a change in outer membrane porin expression, which may reduce permeability.
The isolates identified indicated that there were differences between the contaminated and control-site populations. This was expected, bearing in mind the fact that QACs select for resistant bacteria. QAC disinfectants are inefficient against gram-negative bacteria (1), and this was borne out by the large proportion of gram-negative strains isolated from the reed bed samples and the small number of Bacillus sp. strains isolated compared to those from the agricultural soils. Mesophilic, motile Aeromonas sp. such as A. hydrophila (13 integron-positive isolates) are normal inhabitants of soil and freshwater (13). A study of the effects of QAC-based disinfectants on bacterial community dynamics found an increase in Aeromonas sp. isolates after QAC exposure (24), which agrees with the large number of isolates identified from reed bed samples; aeromonads have also previously been reported to carry class 1 integrons on R plasmids (37). Pseudomonads were common at both sample sites, and as shown in this study, commonly carry class 1 integrons in environments that select for biocide resistance. It is also interesting that many of the genera identified from the contaminated site contain pathogenic species.
It is evident that there are a variety of mechanisms to explain the observed resistance of cultured isolates from the QAC-contaminated reed bed system, but this study is unique in linking increased class 1 integron frequencies with increased QAC resistance. Ninety-five percent of the class 1 integrons carried qacE, reinforcing the hypothesis that QAC pollution selects for integrons. Since class 1 integrons constitute a well-known mechanism for the horizontal transfer of antibiotic resistance genes, it is clear that QAC pollution significantly increases the chances for coselection of antibiotic resistance within environmental bacteria. Recent studies concentrating on investigations of household biocide use and antibacterial resistance (8) failed to establish a link due to the fact that samples were taken from areas where lethal exposures to biocides occur. A comprehensive review of the literature (12) also concluded that the risk of reduced susceptibility to antimicrobial agents associated with exposure to sublethal concentrations of biocides was small. However, these last authors speculated that although biocides at the point of use are rapidly bactericidal and do not therefore pose a risk of increasing antibiotic resistance, downstream of their application a gradient may occur by which sublethal concentrations may provide selective pressure. This hypothesis appears to concur with the data presented in this paper showing that at certain concentrations downstream of application, such as occurs in the reed bed system described here, selection can and does occur for mobile genetic elements carrying biocide resistance genes that are also capable of carrying antibiotic resistance determinants.
We give many thanks to Ronald Skurray for providing reference strains. We also thank Oceans-ESU Ltd. for information regarding the reed bed system.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»