Occurrence of antibiotic resistance and characterization of resistance genes and integrons in Enterobacteriaceae isolated from integrated fish farms in south China

Dynamic Article LinksC<Journal ofEnvironmentalMonitoringCite this: J. Environ. Monit., 2011, 13, 3229

www.rsc.org/jem PAPER

Publ

ishe

d on

05

Oct

ober

201

1. D

ownl

oade

d by

UN

ITE

D S

TA

TE

S G

EO

LO

GIC

AL

SU

RV

EY

on

23/0

9/20

14 1

3:34

:01.

View Article Online / Journal Homepage / Table of Contents for this issue

Occurrence of antibiotic resistance and characterization of resistance genesand integrons in Enterobacteriaceae isolated from integrated fish farms insouth China

Hao-Chang Su,a Guang-Guo Ying,*a Ran Tao,a Rui-Quan Zhang,a Lisa Reynolds Fogartyb

and Dana W. Kolpinc

Received 3rd August 2011, Accepted 7th September 2011

DOI: 10.1039/c1em10634a

Antibiotics are still widely applied in animal husbandry to prevent diseases and used as feed additives to

promote animal growth. This could result in antibiotic resistance to bacteria and antibiotic residues in

animals. In this paper, Enterobacteriaceae isolated from four integrated fish farms in Zhongshan,

South China were tested for antibiotic resistance, tetracycline resistance genes, sulfonamide resistance

genes, and class 1 integrons. The Kirby-Bauer disk diffusion method and polymerase chain reaction

(PCR) assays were carried out to test antibiotic susceptibility and resistance genes, respectively.

Relatively high antibiotic resistance frequencies were found, especially for ampicillin (80%),

tetracycline (52%), and trimethoprim (50%). Out of 203 Enterobacteriaceae isolates, 98.5% were

resistant to one or more antibiotics tested. Multiple antibiotic resistance (MAR) was found highest in

animal manures with a MAR index of 0.56. Tetracycline resistance genes (tet(A), tet(C)) and

sulfonamide resistance genes (sul2) were detected in more than 50% of the isolates. The intI1 gene was

found in 170 isolates (83.7%). Both classic and non-classic class 1 integrons were found. Four genes,

aadA5, aadA22, dfr2, and dfrA17, were detected. To our knowledge, this is the first report for molecular

characterization of antibiotic resistance genes in Enterobacteriaceae isolated from integrated fish farms

in China and the first time that gene cassette array dfrA17-aadA5 has been detected in such fish farms.

Results of this study indicated that fish farms may be a reservoir of highly diverse and abundant

antibiotic resistant genes and gene cassettes. Integrons may play a key role in multiple antibiotic

resistances posing potential health risks to the general public and aquaculture.

aState Key Laboratory of Organic Geochemistry, Guangzhou Institute ofGeochemistry, Chinese Academy of Sciences, Guangzhou, 510640, China.E-mail: [email protected]; Fax: +86 20 85290200; Tel: +86 2085290200; [email protected]. Geological Survey, Lansing, Michigan, USAcU.S. Geological Survey, Iowa City, Iowa, USA

Environmental impact

Integrated fish farming in which animal manures are recycled to feed

Province of South China and in Southeast Asia. The integrated fish f

Animal manure generated from livestock and poultry (commonly p

pipes to fish ponds and the crop fields, which is considered to be

practice may pose potential health risks to the public and aquacult

resistant bacteria harboring abundant varieties of ARGs. Unfor

resistance, particularly resistance genes and integrons in the environ

animal (swine and duck) manure could be a substantial source and

minants. To the best of our knowledge, this is the first report on

Enterobacteriaceae isolated from integrated fish farms.

This journal is ª The Royal Society of Chemistry 2011

Introduction

Antibiotics are not only extensively used for human medicine,

but are also widely used in animal farming for disease therapy

and as feed additives for disease prevention and growth

promotion.1,2 Antibiotic use in the European Union and

freshwater fish in ponds is commonly practiced in Guangdong

arm system usually comprises crops, aquaculture, and livestock.

igs, ducks, and chickens) is directly excreted or transported via

an ecological and economical farming practice. However, this

ure industry, as animal manure could be a significant source of

tunately, little is known about the development of antibiotic

ment related to aquaculture in South China. This study found

reservoir of resistant bacteria carrying diverse resistant deter-

the molecular characterization of antibiotic resistance genes in

J. Environ. Monit., 2011, 13, 3229–3236 | 3229

Fig. 1 Sampling sites of fish farms in Zhongshan, South China. W1–W5

were water samples (W) collected from the Xinchong River (source water

of the fish farms) and the fish farms; D1–D3 were duck manure samples

(D) collected from three duck coops located beside the fish farms; P1 and

P2 were swine manure samples (P) collected from two piggeries located

beside the fish farms; S1–S3 were soil samples (S) collected beside the fish

farms.

Publ

ishe

d on

05

Oct

ober

201

1. D

ownl

oade

d by

UN

ITE

D S

TA

TE

S G

EO

LO

GIC

AL

SU

RV

EY

on

23/0

9/20

14 1

3:34

:01.

View Article Online

Switzerland in 1997 and 1999 were up to 7,659 tons and 8,528

tons, respectively.3 In the USA, the annual usage of antibiotics is

approximately 13,067 tons, with 30% to 70% of this total being

used for livestock.4,5 In China, the annual production of antibi-

otics is 210,000 tons, with 46% being used for animal husbandry.6

The abuse of antibiotics is very common in animal husbandry

and aquaculture for the purposes of feed additives, disease

prevention and growth promotion; and the antibiotics used

include amoxicillin, norfloxacin, gentamicin, neomycin, eryth-

romycin, chloramphenicol, oxytetracycline and clindamycin.6

Among the different classes of antibiotics, tetracyclines and

sulfonamides are the most frequently used veterinary medi-

cines.3,5,7 The widespread human and veterinary use of antibi-

otics and subsequent antibiotic residues in the environment may

well act as a selection pressure for resistant bacteria.2,8 Resistance

genes and integrons may also contribute to the widespread

occurrence and dissemination of antibiotic resistance.9–11

Antibiotic resistance genes (ARGs), as emerging contami-

nants, have received increasing attention.5,12,13 Various ARGs,

such as genes encoding resistance to tetracycline,7,14–17 sulfon-

amides,14–16,18 beta lactams,19 and quaternary ammonium

compounds (qacED1)15 have been widely detected in various

environmental media. Thirty-eight tetracycline resistance genes

have been found in aquatic environments.11 The efflux genes, tet

(A), tet(B), tet(C), tet(D), and tet(E) are frequently detected in

various environmental compartments, including surface water,17

activated sludge of sewage treatment plants,11 fish farms,15 dairy

farms,16 and swine lagoons.11 The sulfonamide resistance genes

sul1, sul2, sul3, and sulA, encoding dihydropteroate synthase,

have been frequently detected in water and sediment in aqua-

culture settings,14,15 surface waters,20 and dairy farms.16

Integrons are mobile genetic elements that can capture, inte-

grate, and express resistance gene cassettes with the help of an

integrase gene.10,11,21,22 Because integrons can contain multiple

resistance genes and have the ability to transfer those genes to

other organisms, they may play an important role in the

dissemination of multiple antibiotic resistance among microor-

ganisms.23 Two kinds of class 1 integrons have been reported.18

Classic class 1 integrons include an integrase gene (intI1) in their

50-conserved segment (50-CS) and the qacED1 and sul1 genes

encoding resistance to quaternary ammonium compounds and

sulfonamides, respectively, in their 30-conserved segment (30-CS).However, non-classic class 1 integrons lack the 30-CS region.18

Several studies have reported the occurrence of integrons/

cassettes in bacteria isolated from clinical samples, food animals,

foods, and rivers.24–29

Because antibiotics are widely used in aquaculture, the pres-

ence of resistant bacteria and ARGs in fish ponds is becoming an

increasing concern. Investigations of antibiotic resistance and

mobile resistant elements in fish/shrimp farms have been carried

out in North America,30,31 South America,32 Europe,33,34 South-

east Asia,35 and Africa.36 Integrated fish farming in which animal

manures are recycled to feed freshwater fish in ponds is

commonly practiced in Guangdong Province of South China37,38

and in Southeast Asia.39 The integrated fish farm system usually

comprises crops, aquaculture, and livestock. Animal manure

generated from livestock and poultry (commonly pigs, ducks,

and chickens) is directly excreted or transported via pipes to fish

ponds and the crop fields, which is considered to be an ecological

3230 | J. Environ. Monit., 2011, 13, 3229–3236

and economical farming practice.39 However, this practice may

pose potential health risks to the public and aquaculture

industry, as animal manure could be a significant source of

resistant bacteria harboring abundant varieties of ARGs.

Unfortunately, little is known about the development of antibi-

otic resistance, particularly resistance genes and integrons in the

environment related to aquaculture in South China.

The objectives of this study were to investigate the occurrence

of bacterial resistance and of tetracycline-resistance genes (tet

(A), tet(B), tet(C), and tet(D)), sulfonamide-resistance genes

(sul1, sul2, and sul3), and class 1 integrons in Enterobacteriaceae

isolated from water, animal manure, and soil samples from four

integrated fish farms in Zhongshan, South China.

Experimental section

Study sites and sample collection

The study sites were located in Zhongshan, South China, which

has four representative integrated fish farms and the Xinchong

River supplying source water for the fish ponds (Fig. 1). Four

kinds of freshwater fishes, blakcarp, grass carp, chub and terch,

which are the most common and most popular freshwater fishes

in China, were raised in the fish ponds. Sampling was carried out

in April (wet season) and December (dry season) 2008 to

compare seasonal differences in bacterial resistance. Water

samples (W1–W5) were collected from the Xinchong River and

four fish ponds, which range in size from about 6000 to 9000 m2

each. The Xinchong River (W1) is the source water for the four

fish farms and is pumped to the four fish ponds when needed.

There is no fish farm upstream from the sampling site W1. The

This journal is ª The Royal Society of Chemistry 2011

Publ

ishe

d on

05

Oct

ober

201

1. D

ownl

oade

d by

UN

ITE

D S

TA

TE

S G

EO

LO

GIC

AL

SU

RV

EY

on

23/0

9/20

14 1

3:34

:01.

View Article Online

integrated fish farms include one duck-fish farm (W2), two duck-

swine-fish farms (W3 andW4) and one swine-fish farm (W5). The

water samples from each fish pond were aseptically collected

from 10 cm below the surface with sterile polyethylene bottles

(500 ml) from three locations (near the shore, pond center, and

opposite from the duck shed or piggery) and composited into

a single sample. Duck manure samples (D: D1, D2, and D3) and

swine manure samples (P: P1 and P2) were collected with

disposable plastic bags in the duck sheds and piggeries in the fish

farms, respectively. Soil samples (S: S1, S2, and S3) were

collected in the vegetable fields beside the fish farms. For each

soil sample, approximately 500 g soil was collected with dispos-

able plastic bags from the surface to 5 cm deep.40 Three samples

were collected from each site, immediately placed on ice, and

transported to the laboratory for further processing.

Isolation of bacteria

Ten-fold serial dilutions (100, 10�1, 10�2, 10�3, 10�4, 10�5 and

10�6) of each water sample were made in sterile saline solution

(0.85% NaCl). Then 0.1ml volumes from each dilution were

spread on nutrient (Oxoid, UK) and MacConkey agar plates

(Oxoid, UK) in triplicate to determine the total culturable

bacteria and Enterobacteriaceae, respectively. The inoculations

were inverted and incubated at 35 �C for 24 h. Colony forming

units (CFUs) on nutrient agar plates were recorded to calculate

the number of total culturable bacteria (CFUs/ml). Meanwhile,

the individual red or pink colonies on MacConkey agar (20–25

isolates per sample) were randomly inoculated onto nutrient agar

plates and enriched at 35 �C for 24 h for further identification.

For solid samples, 5 g of soil or feces were suspended in a 100-ml

sterile saline solution, vibrated vigorously, and left to stand for

one hour. The supernatant was further processed as the water

samples.40

Identification of Enterobacteriaceae

For phenotypic identification of the isolates, the following

biochemical tests were performed: Gram stain, catalase and

oxidase tests, and oxidation-fermentation of glucose. The

isolates that were gram-negative, oxidase-negative, and fermen-

tation-positive, were presumed to be Enterobacteriaceae.41

Escherichia coli ATCC 25922 was used as the control strain.

Antibiotic susceptibility test

Antibiotic susceptibility was tested with the Kirby-Bauer disk

diffusion method on Mueller-Hinton agar plates (Oxoid, UK),

according to the guidelines of the Clinical and Laboratory

Standards Institute.42 A panel of twelve antibiotic discs (Oxoid,

UK) was tested: ampicillin (AMP, 10 mg), piperacillin (PRL, 100

mg), cephazolin (KZ, 30 mg), ceftazidime (CAZ, 30 mg), genta-

micin (CN, 10 mg), streptomycin (S, 10 mg), ciprofloxacin (CIP,

5 mg), levofloxacin (LEV, 5 mg), sulfamethoxazole/trimethoprim

(SXT, 25 mg), trimethoprim (W, 5 mg), tetracycline (TE, 30 mg),

and chloramphenicol (C, 30 mg). The inoculum of Enter-

obacteriaceae was suspended in a sterile saline solution (0.85%

NaCl) with a sterile swab to adjust turbidity to match the 0.5

McFarland standard and streaked evenly on Mueller-Hinton

agar plates. The antibiotic discs were placed on the agar using the

This journal is ª The Royal Society of Chemistry 2011

disc dispenser (Oxoid, UK). The plates were inverted and incu-

bated at 35 �C for 16–18 h. The inhibition zone diameters were

then measured to the nearest millimetre, and the strains were

characterized as susceptible (S), intermediate (I), or resistant (R)

to the antibiotics based on the guidelines of the Clinical and

Laboratory Standards Institute.42 E. coli ATCC 25922 was used

as the control strain.

Identification of tetracycline-resistant and SXT-resistant

Enterobacteriaceae

Presumptive Enterobacteriaceae isolates resistant to TE or SXT

were confirmed with the Enterobacteriaceae identification kit

(Tianhe, China). The Enterobacteriaceae identification kit con-

tained ten biochemical tests, including D-glucose fermentation/

oxidation, L-ornithine decarboxylase, L-lysine decarboxylase,

indole production, H2S production, lactose fermentation/oxida-

tion, galactitol fermentation/oxidation, phenylalanine deami-

nase, urea, and citrate. Purified colonies were inoculated into the

testing tubes and incubated at 36 �C for 18 h. The test results

were recorded and the species of the Enterobacteriaceae were

confirmed according to the handbook supplied with the kit.

Escherichia coli ATCC 25922 was used as the quality control

strain.

PCR assays for detection of resistance genes, integrons and gene

cassettes

Genomic DNA as a template for polymerase chain reaction

(PCR) assays was extracted from the isolates by a boiling tech-

nique.43 A few colonies of each strain were suspended in 100 ml of

distilled water and heated at 100 �C for 20 min, followed

immediately by cooling on ice for 10 min. The suspension was

centrifuged at 16,000 g for 10 min and the supernatant was stored

at �20 �C for PCR assays.43

Multiplex PCR assays were performed to detect the presence

of tetracycline resistance and sulfonamide resistance genes.

Tetracycline-resistant isolates were tested for the presence of

tetracycline resistance genes tet(A), tet(B), tet(C), and tet(D) with

a 25-ml PCR multiplex PCR reaction with: 2 ml template DNA, 1

� PCR buffer (TaKaRa, Japan), 0.3 mM dNTP mix (TaKaRa,

Japan), 4 mM MgCl2 (TaKaRa, Japan), 1.25 units Taq poly-

merase (TaKaRa, Japan), 0.05mM each primer (Table 1), 0.25ml

dimethyl sulfoxide, and distilled water. A second PCR multiplex

was performed on sulfonamide-resistant strains for sulfonamide

resistance genes, sul1, sul2, sul3. These PCR assays were carried

out in a 25-ml reaction mixture, which consisted of 3 ml of

template DNA, 1 � PCR buffer (TaKaRa, Japan), 0.3 mM

dNTP mix (TaKaRa, Japan), 2 mM MgCl2 (TaKaRa, Japan),

1.5 units Taq polymerase (TaKaRa, Japan), 0.1 mM each primer

(Table 1) and distilled water. The temperature program for

detection of tet genes and sul genes consisted of initial denaturing

at 94 �C for 5 min, followed by 30 cycles of 30 s at 94 �C; 30 s at

the annealing temperature (55 �C for tet genes, 65 �C for sul

genes); 60 s at 72 �C, and a final extension step for 10 min at

72 �C.12,44,45 The PCR products were subjected to gel electro-

phoresis. Amplified DNA (5 ml) was mixed with 1 ml 6� loading

buffer dye and loaded on 2% (w/v) agarose gel containing 1 �Gelred nucleic acid stain (Biotium, USA), followed by running in

J. Environ. Monit., 2011, 13, 3229–3236 | 3231

Table 1 PCR primers targeting tetracycline resistance genes, sulfonamide resistance genes, and Class 1 integrons

Targeted gene class Sequence (50 to 30) Annealing T/�C Amplicon size (bp) Reference

Class 1 integron 50-CS GGC ATC CAA GCA GCA AG 55 variable 2130-CS AAG CAG ACT TGA CCT GA

aadA1 or aadA2 aadA-RVa ATC CTT CGG CGC GAT TTT G 55 variable 55intI1 FW CAG TGG ACA TAA GCC TGT TC 55 160 56

RV CCC GAG GCA TAG ACT GTAtet(A) FW GCT ACA TCC TGC TTG CCT TC 60 210 45

RV CAT AGA TCG CCG TGA AGA GGtet(B) FW TTG GTT AGG GGC AAG TTT TG 58 659 45

RV GTA ATG GGC CAA TAA CAC CGtet(C) FW CTT GAG AGC CTT CAA CCC AG 60 418 45

RV ATG GTC GTC ATC TAC CTG CCtet(D) FW AAA CCA TTA CGG CAT TCT GC 60 787 45

RV GAC CGG ATA CAC CAT CCA TCsul1 FW CGG CGT GGG CTA CCT GAA CG 56 433 44

RV GCC GAT CGC GTG AAG TTC CGsul2 FW GCG CTC AAG GCA GAT GGC ATT 60.8 293 44

RV GCG TTT GAT ACC GGC ACC CGTsul3 FW TCCGTT CAGCGAATTGGTGCAG 60 128 12

RV TTC GTT CAC GCC TTA CAC CAG C

a FW, forward; RV, reverse.

Publ

ishe

d on

05

Oct

ober

201

1. D

ownl

oade

d by

UN

ITE

D S

TA

TE

S G

EO

LO

GIC

AL

SU

RV

EY

on

23/0

9/20

14 1

3:34

:01.

View Article Online

1 � TAE buffer at 5V cm�1 for 40 min, and visualizing by

ultraviolet (UV) transillumination.

Integron PCR assays were performed to detect the class 1

integrons and gene cassettes in the isolates. The PCR assays were

carried out in a 25-ml reaction mixture containing: 5 ml template

DNA, 1 � PCR buffer (TaKaRa, Japan), 0.2 mM dNTP mix

(TaKaRa, Japan), 1.5 mM MgCl2 (TaKaRa, Japan), 2.5 units

Taq polymerase (TaKaRa, Japan), 0.4 mM each primer, and

distilled water. For detection of class 1 integrase (intI1), primers

intI1-FW and intI1-RV were used. For detection of class 1

integrons, primers specific for the 50-conserved segment and

30-conserved segment of class 1 integrons were used. Primers

50-CS and aadA-RV were used to detect the gene cassettes of

class 1 integrons when intI1 PCRwas positive and no sul1 gene or

no PCR amplicon of the class 1 integron variable region was

detected (Table 1). Amplification was carried out by denaturing

for 5 min at 94 �C, followed by 35 cycles of 94 �C for 1 min, 55 �Cfor 1 min, 72 �C for 5 min, and a final extension step for 10 min at

72 �C. Subsequent PCR product was subjected to sequencing

(BGI, China) and nucleotide sequence comparison was per-

formed by the Basic Local Alignment Search Tool (BLAST)

through the National Center for Biotechnology Information

(NCBI).

PCR assays were run on Bio-Rad S1000 Thermal Cycler (Bio-

Rad, USA). Each PCR run contained positive controls, a nega-

tive control (DNA extraction of E.coli ATCC 25922), and

a blank control (distilled water instead of DNA extraction). The

following E.coli strains, which were isolated in our laboratory,

served as positive controls: strains S1-10 for tet(A), H10-21 for

tet(B), B0-31 for tet(C) and H10-5 for tet(D) detection; strains

H12-1-5 for sul1 and sul2 and H3-2-10 for sul3 detection; and

strains DG3-15, DC2-5, and YC1-15 for class 1 integrase and

gene cassettes detection. Each PCR assay was run in duplicate.

Statistical analysis

The antibiotic resistance frequency was calculated by the equa-

tion: m/n � 100%, where m is the number of Enterobacteriaceae

3232 | J. Environ. Monit., 2011, 13, 3229–3236

resistant to antibiotic, and n is the number of Enterobacteriaceae

isolated from the sample.

The multiple antibiotic resistance (MAR, resistant to 2 or

more antibiotics) index was determined for each sample site by

the equation: a/(b � c), where a is the aggregate resistance score

of all isolates from the sample, b is the number of antibiotics

tested, and c is the number of Enterobacteriaceae isolated from

the sample.46,47

To explore the similarity of antibiotic resistance of the samples

contaminated by animal manures, hierarchical cluster analysis of

bacterial resistance frequency data from every site was performed

to classify the samples by using the minimum distance method.

Statistical tests were performed using Data Processing System

9.50 (Refine Information Tech. Co., LtD, Hangzhou, China).

Duncan’s multiple range test was performed to evaluate the

statistical significance of difference with p < 0.05. Averages and

standard deviations were calculated with Microsoft Excel, 2003.

Results

Total culturable bacteria and prevalence of antibiotic resistance

As shown in Table 2, the total culturable bacteria numbers were

significantly lower in the water samples from the Xinchong River

(source water) compared to water samples from the fish ponds,

whereas the bacterial numbers in the manure samples were

significantly higher than in the soil samples. Out of 203 Enter-

obacteriaceae strains isolated from fish farm samples collected in

April and December 2008 in Zhongshan, South China, 200

(98.5%) were resistant to one or more of 12 antibiotics tested.

Multiple antibiotic resistance was frequently detected, with 77 of

the isolates (37.9%) resistant to seven antibiotics, more than half

of the antibiotics tested. Two isolates, identified as E. coli, were

resistant to all twelve antibiotics tested. These isolates were

obtained from a duck manure sample.

The frequencies of antibiotic resistance for Enterobacteriaceae

isolated from the fish farms are shown in Table 2 and Fig. 2. The

Enterobacteriaceae isolates appeared resistant in various degrees

This journal is ª The Royal Society of Chemistry 2011

Table 2 Frequencies of antibiotic resistance among Enterobacteriaceae isolated from fish farm samples

Samples (No. of isolates) CFUs/ml; CFUs/g(mean � SD) a

Frequency of resistant Enterobacteriaceae (%) b

MAR index cAMP PRL KZ CAZ CN S CIP LEV SXT W TE C

April 2008 W1 (12) (3.3 � 0.5) � 103 43 9 17 7 18 21 0 0 19 24 24 15 0.12 � 0.02cW2 (12) (3.6 � 0.5) � 104 77 23 32 14 24 34 10 10 37 32 45 21 0.23 � 0.02bW3 (13) (4.1 � 0.8) � 104 89 22 33 21 33 35 14 14 43 45 56 30 0.34 � 0.02bW4 (13) (4.3 � 0.8) � 104 91 28 35 20 35 37 15 15 47 53 58 29 0.31 � 0.04bW5 (13) (4.4 � 0.7) � 104 90 25 37 30 33 36 14 14 52 51 60 33 0.31 � 0.03bD (13) (3.5 � 0.7) � 1013 100 36 47 36 45 53 19 19 59 69 67 31 0.53 � 0.05aP (13) (4.2 � 0.7) � 1013 93 34 44 38 42 57 20 20 63 75 75 47 0.56 � 0.05aS (12) (1.0 � 0.2) � 108 65 12 20 13 20 29 9 9 23 37 34 50 0.24 � 0.02b

December 2008 W1 (12) (2.9 � 0.6) � 103 45 10 15 9 14 20 0 0 20 28 21 17 0.13 � 0.03cW2 (13) (3.3 � 0.6) � 104 69 22 33 12 21 32 10 10 33 39 40 26 0.20 � 0.02bW3 (13) (3.8 � 0.7) � 104 85 23 34 20 33 33 13 13 39 49 47 33 0.31 � 0.03bW4 (13) (4.0 � 0.7) � 104 88 26 37 22 34 38 14 14 30 51 51 34 0.33 � 0.01bW5 (13) (4.3 � 0.6) � 104 92 24 34 28 31 36 13 13 54 48 66 29 0.31 � 0.03bD (13) (3.2 � 0.6) � 1013 96 33 48 36 44 57 20 20 69 71 78 34 0.55 � 0.08aP (13) (4.9 � 0.9) � 1013 93 35 47 35 45 59 21 21 67 82 77 51 0.54 � 0.04aS (13) (9.6 � 0.2) � 107 67 11 21 15 21 28 8 8 26 38 31 53 0.22 � 0.03b

All samples (203) 80 23 33 22 31 38 13 13 43 50 52 33 0.33

a CFUs/ml: colony forming units per millilitre of total culturable bacteria in water samples, river samples (W1), duck-fish pond samples (W2), duck-swine-fish pond samples (W3 and W4), swine-fish pond (W5); CFUs/g: colony forming units per gram of total culturable bacteria in manure andsoil samples, duck manure (D), swine manure (P) and soil samples (S). b AMP: ampicillin; PRL: piperacillin; KZ: cephazolin; CAZ: ceftazidime;CN: gentamicin; S: streptomycin; CIP: ciprofloxacin; LEV: levofloxacin; SXT: sulfamethoxazole/trimethoprim; W: trimethoprim; TE: tetracycline;C: chloramphenicol. c MAR index: multiple antibiotic resistance index (mean � SD) (n ¼ 3). abc: signicant difference indicated by different letters,Duncan’s multiple range test, p < 0.05.

Publ

ishe

d on

05

Oct

ober

201

1. D

ownl

oade

d by

UN

ITE

D S

TA

TE

S G

EO

LO

GIC

AL

SU

RV

EY

on

23/0

9/20

14 1

3:34

:01.

View Article Online

to all 12 antibiotics tested. The resistance frequency was noted

with ampicillin having the greatest resistance frequency and

ciprofloxacin and levofloxacin having the lowest resistance

frequency (Fig. 2).

The multiple antibiotic resistance (MAR) indexes are pre-

sented in Table 2. The river samples had the lowest MAR index

(0.12), and the swine manure samples had the highest MAR

index (0.56). The MAR indexes for animal (swine and duck)

manure were significantly higher than those of soil samples and

water samples in the fish farms. The pond water and soil samples

did not have statistically different MAR values. No seasonal

difference was found for the resistance frequency and MAR

Fig. 2 Total frequency of Enterobacteriaceae isolates resistant to the 12

antibiotics tested. AMP: ampicillin; PRL: piperacillin; KZ: cephazolin;

CAZ: ceftazidime; CN: gentamicin; S: streptomycin; CIP: ciprofloxacin;

LEV: levofloxacin; SXT: sulfamethoxazole/trimethoprim; W: trimetho-

prim; TE: tetracycline; C: chloramphenicol. Letters a, b, and c: statisti-

cally signicant difference indicated by different letters, with Duncan’s

multiple range test, p < 0.05. Error bars are standard deviations of the

resistance frequencies of 203 isolates.

This journal is ª The Royal Society of Chemistry 2011

index, suggesting no change in river water quality in terms of

bacterial resistance between the wet and dry seasons.

Detection of tetracycline and sulfonamide resistance genes

All four tet genes (tet(A), tet(B), tet(C), and tet(D)) and three sul

genes (sul1, sul2, and sul3)were detected in the Enter-

obacteriaceae isolates in the present study. In 105 tetracycline-

resistant Enterobacteriaceae isolates, tet(A) and tet(C) were the

two most prevalent genes, with the detection frequencies of

74.3% and 62.9%, respectively, followed by tet(B) (23.8%), and

tet(D) (10.5%). Among the 12 gene patterns observed (Table 3),

the most frequent pattern was tet(A) detected in 27 isolates

(25.7%). The combination of tet(C) + tet(D) and tet(A) + tet(B) +

tet(C) + tet(D) were both found in only one isolate (1.0%).

Frequencies of the other nine patterns ranged from 1.9% to

23.8%. No tet(B) + tet(D) or tet(A) + tet(B) + tet(D) pattern was

found. The results indicated that tet genes were primarily

detected in Escherichia coli (77.1%). Other Enterobacteriaceae

isolates identified carrying a tet gene included: Enterobacter

aerogenes (6.3%), Salmonella arizonae (5.3%), Salmonella spp.

(5.5%), and Serratia marcescens (5.8%).

Out of the 203 Enterobacteriaceae isolates, 87 isolates were

resistant to sulfamethoxazole/trimethoprim (SXT), and 66 of

those isolates contained a sulfonamide resistance (sul) gene of the

three sul genes (sul1, sul2, and sul3). The sul2 gene was the most

frequently detected sul gene (89.4% out of 66 isolates), followed

by sul1 (50%), and sul3 (3%). In addition, five gene patterns were

observed (Table 3). Among 66 sul gene positive isolates, gene

pattern sul2 was the most prevalent, with a frequency of 48.5%,

followed by the pattern both sul1 and sul2 (39.4%) detected in 26

of the isolates. In one isolate sul1, sul2, and sul3 were all detected.

J. Environ. Monit., 2011, 13, 3229–3236 | 3233

Table 3 Prevalence of resistance genes, integrons, and gene cassettes inEnterobacteriaceae isolated from fish farm samples

Resistant genes,integrons and genecassettes

No. ofEnterobacteriaceaeisolates (%)

Strains distribution by sampletype (No. of strains)

tet genestet A 27 (25.7) River (5), fish ponds (19), duck

manure (37), swine manure (28),soil (16)

tet B 2 (1.9)tet C 18 (17.1)tet D 3 (2.9)tet A/B 3 (2.9)tet A C�1 25 (23.8)tet A/D 4 (3.8)tet B/C 3 (2.9)tet C D�1 1 (1.0)tet A/B/C 16 (15.2)tet A C�1 D�1 2 (1.9)tet A/B/C D�1 1 (1.0)Total 105 (100)

sul genessul1 6 (9.1) River (3), fish ponds (13), duck

manure (20), swine manure (18),soil (12)

sul2 32 (48.5)sul3 1 (1.5)sul1 + sul2 26 (39.4)sul1 + sul2 + sul3 1 (1.5)Total 66 (100)

Integrons and gene cassettesIntI1-positive 170 (83.7) River (3), fish ponds (21), duck

manure (34), swine manure (26),soil (16)

With gene cassettes 100 (49.3)aadA22 31 (31)dfr2 7 (7.0)dfrA17-aadA5 29 (29)aadA22 + dfr2 17 (17)aadA22 + dfrA17-aadA5

10 (10)

aadA22 + dfr2 +dfrA17-aadA5

6 (6.0)

Publ

ishe

d on

05

Oct

ober

201

1. D

ownl

oade

d by

UN

ITE

D S

TA

TE

S G

EO

LO

GIC

AL

SU

RV

EY

on

23/0

9/20

14 1

3:34

:01.

View Article Online

E. coli represented 75.8% of the isolates in which a sul gene was

detected (50/66). Other Enterobacteriaceae isolates that con-

tained a sul gene included: Enterobacter aerogenes (9.1%),

Salmonella arizonae (4.5%), Salmonella spp. (4.5%), and Serratia

marcescens (6.1%).

Class 1 integron and gene cassettes analysis

Overall, 170 Enterobacteriaceae isolates (83.7%) contained class

1 integrons (Table 3). The intI1 gene was detected in 117 isolates

(68.8%) of Escherichia coli, 28 isolates (16.5%) of Enterobacter

aerogenes, 10 isolates (5.9%) of Serratia marcescens, nine isolates

(5.3%) of Salmonella spp. and six isolates (3.5%) of Salmonella

arizonae. Of the 170 intI1-positive isolates, 100 isolates (58.8%)

harbored gene cassette(s). Among them, 33 isolates (33%) were

classic integrons; whereas unexpectedly, the other 67 isolates

(67%) were non-classic integrons. Three gene cassettes were

detected: aadA22, the gene cassette encoding aminoglycoside

adenyltransferase, conferring resistance to streptomycin and

spectinomycin; dfrA17-aadA5, encoding dihydrofolate reductase

and aminoglycoside adenylyltransferase; and dfr2, encoding

trimethoprim dihydrofolate reductase (Table 3). It is noted that

six different gene cassette arrays were identified as follows:

aadA22 in 31 isolates (31.0%); dfrA17-aadA5 in 29 isolates

3234 | J. Environ. Monit., 2011, 13, 3229–3236

(29.0%); aadA22 + dfr2 in 17 isolates (17.0%); aadA22 + dfrA17-

aadA5 in 10 isolates (10.0%); dfr2 in 7 isolates (7.0%) and aadA22

+ dfr2 + dfrA17-aadA5 in 6 isolates (6.0%). These different arrays

and associated sulfonamide and trimethoprim resistance and sul

gene detection are presented in Table 4.

Discussion

High prevalence of antibiotic resistance and multiple resistances

were observed in this study. Out of 203 Enterobacteriaceae

strains isolated from four representative integrated fish farms in

Zhongshan, South China, a high frequency of isolates (98.5%)

displayed resistance to one or more antibiotics tested. Mean-

while, 37.9% were multiple resistances, resistant to seven of the

12 antibiotics tested, with two E.coli strains isolated from duck

manure resistant to all twelve antibiotics tested. Animal manures

are proposed as the source of antibiotic resistance bacteria based

on the results of bacterial phenotype and genotype for the four

integrated fish farms.

The highest frequency of the 12 antibiotics tested was found in

duck and swine manure samples (Table 2). No resistance,

however, was documented in the Xinchong River (source water)

samples for the two fluoroquinolones tested (ciprofloxacin and

levofloxacin) which are primarily human health medicines. Flu-

oroquinolone resistance in the animal manure samples could be

explained by illegal use of these antibiotics in swine, ducks or fish

feed.48 Nevertheless, compared to the other antibiotics, the

resistance to the two fluoroquinolones was much lower than the

other target antibiotics.

The MAR index could provide useful information to evaluate

the health risk of bacterial resistance.46,47 The order of hazard

based on the MAR indexes would be: animal manure (0.55) >

fish farm water and soil (0.28) > river source water (0.13) as

shown in Table 2. Meanwhile, hierarchical cluster of resistance

frequencies data was identical with the fish farm conditions

(Fig. 3). Animal manures (D and P) had the same antibiotic

resistance profile. Fish ponds W3 and W4 were the same eco-fish

culturing pattern (ducks, swine and fish). W2 and W5 were both

a sole-animal culturing pattern. Fish pond W2 was close to

a duck shed (D) whereas fish pondW5 was close to a piggery (P),

and both fish ponds received input of duck manure and swine

manure, respectively. In addition, S (soil) was grouped with fish

ponds and animal manures, suggesting that direct excretion or

discharge of animal manures to fish ponds and crop fields

increases antibiotic resistance levels of the receiving environ-

ments. As expected, the source water (W1) of the fish farms was

clustered far from the fish farms group, indicating that animal

manures and fish farms could be the origin of the antibiotic

resistance among Zhongshan fish farm environment settings.

Bacterial resistance to antibiotics is mainly due to harboring of

resistance determinants in bacteria.11 In the present study, four

tetracycline efflux genes (tet(A), tet(B), tet(C), and tet(D)), three

sulfonamide resistance genes (sul1, sul2, and sul3) and class

1 gene cassettes were detected in Enterobacteriaceae isolates. The

presence of tet genes has previously been reported in fish farms

and marine sediments.14,30,49 Sulfonamide resistance gene sul2

was widely detected in the strains isolated from river water, fish

farm water, animal manure, and soil samples, with a frequency of

67.8% out of 87 SXT-resistant strains. Kerrn et al.44 and Peirano

This journal is ª The Royal Society of Chemistry 2011

Table 4 Gene cassette(s) of class 1 integron and related antibiotic resistance gene and resistance patterns in Enteroacteriaceae isolated from fish farmsamples

intI1 gene Gene cassette array Strains carrying gene cassette(s) No. of isolates Accession No.

sul genesAntibioticresistancea

sul1 sul2 sul3 SXT W STR

+ dfr2 Escherichia coli 7 JF412519 + R R S+ aadA22 Escherichia coli 24 JF412520 + R R R

Enterobacter aerogenes 2 + R R RSalmonella arizonae 2 + R R RSalmonella spp. 1 + R R RSerratia marcescens 2 + R R R

+ dfrA17-aadA5 Escherichia coli 19 JF412521 + R R REnterobacter aerogenes 10 + R R R

+ aadA22 + dfr2 Escherichia coli 17 JF412519 + R R RJF412520

+ aadA22 + dfrA17-aadA5 Escherichia coli 7 JF412520 + + R R REnterobacter aerogenes 2 JF412521 + R R RSerratia marcescens 1 + R R R

+ daadA22 + dfr2 + dfrA17-aadA5 Escherichia coli 6 JF412519 + + R R RJF412520JF412521

a SXT: sulfamethoxazole/trimethoprim; W: trimethoprim; STR: streptomycin; +: gene detected; R: resistant; S: susceptible.

Publ

ishe

d on

05

Oct

ober

201

1. D

ownl

oade

d by

UN

ITE

D S

TA

TE

S G

EO

LO

GIC

AL

SU

RV

EY

on

23/0

9/20

14 1

3:34

:01.

View Article Online

et al.50 also reported that sul2 was prevalently found among

Enterobacteriaceae. In the present study, among the species of

Enterobacteriaceae identified (203 isolates), we found that

frequency of sul2 was dominant in E.coli (29.1%), followed by

gene sul1 (16.3%), and sul3 (1%). It is interesting to note that sul1,

sul2, or sul3 were not detected in 21 out of 87 SXT-resistant

isolates (24.1%). These isolates may harbor other genes such as

the sulA resistance determinant or dfr group genes encoding

dihydrofolate reductase.11

Integrons are of increasing concern globally as they have the

ability to capture different kinds of resistance gene cassettes

resulting in multiple antibiotic resistances.10 Therefore, integrons

play a key role in spreading antibiotic resistance genes in

microorganisms. In this study, the detection frequency of class 1

integrase (intI1) (83.7%) is much higher than 6.9%, reported by

Ishida et al.36 It is noted that both classic class 1 integrons and

non-classic class 1 integrons were detected in this study, with

frequencies of 33.0% and 67.0%, respectively. Previous research

has also found the sulfonamide resistance gene sul1 was associ-

ated with the classic class 1 integron.18 Four gene cassettes

Fig. 3 Hierarchical cluster analysis based on the bacterial resistance

frequencies data. W1–W5: water samples collected from the Xinchong

River and the fish ponds; D: duck manure smples; P: swine manure

samples; S: soil samples. With the cluster method of between-groups

linkage, significance value is at p < 0.05.

This journal is ª The Royal Society of Chemistry 2011

(aadA5, aadA22, dfr2, and dfrA17) and six cassette patterns were

observed in this study. The high prevalence and widespread

occurrence of integrons and gene cassettes in Enterobacteriaceae

isolated from the river water, fish farm water, animal manure,

and soil samples may constitute a public health hazard, as

multiple gene cassettes could be captured and integrated result-

ing in multiple antibiotic resistance, even to broad-spectrum

antibiotics such as the third-generation cephalosporins and

fluoroquinolones.24

Gene cassettes of the aadA and dfr groups have been widely

found in sewage effluents, natural water, sediments and hospital

wastewaters, and in animal production and aquaculture

areas.11,51 The aadA group of genes encodes resistance to strep-

tomycin and spectinomycin, whereas the dfr group encodes

resistance to trimethoprim. In the present study, 50% of the

Enterobacteriaceae isolates from the fish farm samples were

resistant to trimethoprim and 43% resistant to sulfamethoxazole/

trimethoprim (SXT). Consistent with the resistance findings, sul

and dfr genes were detected in some of 203 isolates from the fish

farm samples: sul1 (16.3%), sul2 (29.1%), sul3 (1%), dfr2 (14.8%)

and dfrA17 (22.2%). Cernat et al.51 also reported a higher prev-

alence for dfrA17 (70%), dfrA12 (50%), sul1 (50%) and sul2 (60%)

in E. coli strains isolated from drinking, recreational, and sal-

master waters (marine waters) in Romania. In addition, study

results documented that dfrA17-aadA5, first described by White

et al.,52 was the most common gene cassette array, conferring

high level resistance to trimethoprim and spectinomycin. To our

knowledge, this is the first documentation of gene cassette array

dfrA17-aadA5 detected in integrated fish farms.

Among five sample types (Table 3), tet genes, sul genes, and

class 1 gene cassettes were detected most often in animal manures

(61.9%, 57.6% and 60.0%, respectively), followed by fish ponds

(18.1%, 19.7% and 21.0%, respectively), soil samples (15.2%,

18.2% and 16.0%, respectively), and river water (4.8%, 4.5% and

3.0%, respectively). In addition, among all the genes analyzed, tet

(D), sul3, and six E. coli isolates containing the four-gene cassette

J. Environ. Monit., 2011, 13, 3229–3236 | 3235

Publ

ishe

d on

05

Oct

ober

201

1. D

ownl

oade

d by

UN

ITE

D S

TA

TE

S G

EO

LO

GIC

AL

SU

RV

EY

on

23/0

9/20

14 1

3:34

:01.

View Article Online

combination of aadA22 + dfr2 + dfrA17-aadA5 were only

detected in animal manure. The results suggest that animal

(swine and duck) manure may be a substantial source and

possible reservoir of resistant bacteria carrying diverse resistant

determinants. Several studies have also identified antibiotic

resistant organisms and genes associated with manure.2,11,53,54

The fate and transfer of these organisms and genes in the envi-

ronment, however, are less understood. Cluster analysis of the

frequencies of antibiotic resistance does indicate that those

waters which receive manure input are more closely related to the

manure resistance frequencies than those frequencies found in

the river water and nearby soil samples, again suggesting animal

manure is a likely source of resistance in these fish farm systems.

Acknowledgements

The authors would like to acknowledge the support from the

CAS Key Project (KZCX2-EW-108) and National Natural

Science Foundation of China (NSFC 40688001, 40821003 and

40771180) and Guangdong Provincial Science Foundation

(8251064004000001). Any use of trade, firm, or product names is

for descriptive purposes only and does not imply endorsement by

the authors. This is a contribution No. 1391 from GIG CAS.

References

1 B. Halling-Sorensen, S. N. Nielsen, P. F. Lanzky, F. Ingerslev,H. C. Holten Lutzhoft and S. E. Jorgensen, Chemosphere, 1998, 36,357–394.

2 J. Luis Martinez, Environ. Pollut., 2009, 157, 2893–2902.3 F. R. Ungemach, D. Mueller-Bahrdt and G. Abraham, Int. J. Med.Microbiol., 2006, 296, 33–38.

4 FDA, U.S. Food and Drug Administration, Department of Healthand Human Services, 2009.

5 C. W. McKinney, K. A. Loftin, M. T. Meyer, J. G. Davis andA. Pruden, Environ. Sci. Technol., 2010, 44, 6102–6109.

6 Y.-P. Wang and Y. Ma, Chin. J. Antibiotics, 2008, 33, 519–523.7 A. J. Patterson, R. Colangeli, P. Spigaglia and K. P. Scott, Environ.Microbiol., 2007, 9, 703–715.

8 H. A. Duong, N. H. Pham, H. T. Nguyen, T. T. Hoang, H. V. Pham,V. C. Pham, M. Berg, W. Giger and A. C. Alder, Chemosphere, 2008,72, 968–973.

9 F. Baquero, J. L. Martinez and R. Canton, Curr. Opin. Biotechnol.,2008, 19, 260–265.

10 J. A. Di Conza and G. O. Gutkind, Rev. Argent. Microbiol., 2010, 42,63–78.

11 M. C. Roberts, FEMS Microbiol. Lett., 2005, 245, 195–203.12 R. T. Pei, S. C. Kim, K. H. Carlson and A. Pruden,Water Res., 2006,

40, 2427–2435.13 A. Pruden, R. T. Pei, H. Storteboom and K. H. Carlson, Environ. Sci.

Technol., 2006, 40, 7445–7450.14 Y. Agerso and A. Petersen, J. Antimicrob. Chemother., 2006, 59, 23–

27.15 O. Akinbowale, H. Peng and M. Barton, Int. J. Antimicrob. Agents,

2007, 29, S113–S113.16 V. Srinivasan, H. M. Nam, L. T. Nguyen, B. Tamilselvam,

S. E. Murinda and S. P. Oliver, Foodborne Pathog. Dis., 2005, 2,201–211.

17 R. Tao, G. G. Ying, H. C. Su, H.W. Zhou and J. P. S. Sidhu, Environ.Pollut., 2010, 158, 2101–2109.

18 L. Vinue, Y. Saenz, B. Rojo-Bezares, I. Olarte, E. Undabeitia,S. Somalo, M. Zarazaga and C. Torres, Int. J. Antimicrob. Agents,2010, 35, 492–496.

19 E. Cendejas, R. Gomez-Gil, P. Gomez-Sanchez and J. Mingorance,Clin. Microbiol. Infect., 2010, 16, 181–183.

20 J. Y. Hu, J. C. Shi, H. Chang, D. Li, M. Yang and Y. C. Kamagata,Environ. Sci. Technol., 2008, 42, 3415–3420.

3236 | J. Environ. Monit., 2011, 13, 3229–3236

21 C. Levesque, L. Piche, C. Larose and P. H. Roy, Antimicrob. AgentsChemother., 1995, 39, 185–191.

22 H. W. Stokes and R. M. Hall, Mol. Microbiol., 1989, 3, 1669–1683.23 A. M. Ahmed, Y. Motoi, M. Sato, M. Maruyama, H. Watanabe,

Y. Fukumoto and T. Shimamoto, Appl. Environ. Microbiol., 2007,73, 6686–6690.

24 H. Chen, W. Q. Shu, X. S. Chang, J. A. Chen, Y. B. Guo and Y. Tan,Environ. Pollut., 2010, 158, 2459–2464.

25 M. S. Kang, A. Kim, B. Y. Jung, M. Her, W. Jeong, Y. M. Cho,J. Y. Oh, Y. J. Lee, J. H. Kwon and Y. K. Kwon, Avian Pathol.,2010, 39, 201–205.

26 Y. Khosravi, S. T. Tay and V. Jamuna, Int. J. Infect. Dis., 2010, 14,e37.

27 L. Lapierre, B. S. Martin, C. Araya-Jordan and C. Borie, Can. J.Microbiol., 2010, 56, 515–521.

28 L. Song, Y. B. Ning, J. Z. Shen, X. Z. Fan, C. P. Zhang, C. H. Yangand J. F. Han, Sci. China-Life Sci., 2010, 53, 613–619.

29 Y. C. Wang, Y. C. Chang, H. L. Chuang, C. C. Chiu, K. S. Yeh,C. C. Chang, S. L. Hsuan and T. H. Chen, Afr. J. Microbiol. Res.,2010, 4, 677–681.

30 E. E. Seyfried, R. J. Newton, K. F. Rubert, J. A. Pedersen andK. D. McMahon, Microb. Ecol., 2010, 59, 799–807.

31 M. Uyaguari, P. Key, J. Moore, K. Jackson and G. Scott, Environ.Toxicol. Chem., 2009, 28, 2715–2724.

32 C. D. Miranda, C. Kehrenberg, C. Ulep, S. Schwarz andM. C. Roberts, Antimicrob. Agents Chemother., 2003, 47, 883–888.

33 H. Le Bris, R. Dhaouadi, M. Naviner, E. Giraud, C. Mangion,F. O. Armand, N. De la Cotte, C. Thorin, J. P. Ganiere andH. Pouliquen, Aquaculture, 2007, 273, 416–422.

34 A. S. Schmidt, M. S. Bruun, I. Dalsgaard, K. Pedersen andJ. L. Larsen, Appl. Environ. Microbiol., 2000, 66, 4908–4915.

35 A. Petersen, J. S. Andersen, T. Kaewmak, T. Somsiri andA. Dalsgaard, Appl. Environ. Microbiol., 2002, 68, 6036–6042.

36 Y. Ishida, A. M. Ahmed, N. B. Mahfouz, T. Kimura, S. A. El-Khodery, A. A. Moawad and T. Shimamoto, J. Vet. Med. Sci.,2010, 72, 727–734.

37 Y. T. Yang, Guangdong Agricultural Sciences, 1995, 5, 14–16.38 X. B. Wang and Q. S. Hu, Hunan Siliao, 2009, 4, 22.39 O. Mikolasek, T. D. Khuyen, J. M. Medoc and V. Porphyre, Cah.

Agric., 2009, 18, 235–241.40 L. B. Jensen, S. Baloda, M. Boye and F. M. Aarestrup, Environ. Int.,

2001, 26, 581–587.41 R. E. Buchanan and N. E. Gibbons, The Williams & Wilkins

Company, Baltimore, 1974.42 CLSI, Clinical and Laboratory Standards Institute, Wayne, PA., 2009.43 M. Gueimonde, S. Tolkko, T. Korpimaki and S. Salminen, Appl.

Environ. Microbiol., 2004, 70, 4165–4169.44 M. B. Kerrn, T. Klemmensen, N. Frimodt-Moller and F. Espersen, J.

Antimicrob. Chemother., 2002, 50, 513–516.45 L. K. Ng, I. Martin, M. Alfa andM.Mulvey,Mol. Cell. Probes, 2001,

15, 209–215.46 P. H. Krumperman, Appl. Environ. Microbiol., 1983, 46, 165–170.47 C. D. Miranda and R. Zemelman, Sci. Total Environ., 2002, 293, 207–

218.48 J. F. Yang, G. G. Ying, L. J. Zhou, S. Liu and J. L. Zhao, Environ.

Pollut., 2009, 157, 2704–2709.49 M. H. Rahman, L. Nonaka, R. Tago and S. Suzuki, Environ. Sci.

Technol., 2008, 42, 5055–5061.50 G. Peirano, Y. Agerso, F. M. Aarestrup and D. D. Rodrigues, J.

Antimicrob. Chemother., 2005, 55, 301–305.51 R. Cernat, C. Balotescu, D. Ivanescu, D. Nedelcu, V. Lazar,

M. Bucur, D. Valeanu, R. Tudorache, M. Mitache andM. Dragoescu, Int. J. Antimicrob. Agents, 2007, 29, S274–S274.

52 P. A. White, C. J. McIver, Y. M. Deng and W. D. Rawlinson, FEMSMicrobiol. Lett., 2000, 182, 265–269.

53 C. A. Engemann, P. L. Keen, C. W. Knapp, K. J. Hall andD. W. Graham, Environ. Sci. Technol., 2008, 42, 5131–5136.

54 H. Storteboom, M. Arabi, J. G. Davis, B. Crimi and A. Pruden,Environ. Sci. Technol., 2010, 44, 7397–7404.

55 Y. Saenz, L. Brinas, E. Dominguez, J. Ruiz, M. Zarazaga, J. Vila andC. Torres, Antimicrob. Agents Chemother., 2004, 48, 3996–4001.

56 J. G. M. Koeleman, J. Stoof, M. W. Van der Bijl, C. Vandenbroucke-Grauls and P. H. M. Savelkoul, J. Clin. Microbiol., 2001, 39, 8–13.

This journal is ª The Royal Society of Chemistry 2011