1
Resistance to broad-spectrum antibiotics in aquatic systems: anthropogenic activities 1
modulate the dissemination of blaCTX-M-like genes. 2
Marta Tacão, António Correia, and Isabel Henriques 3
Biology Department and CESAM, University of Aveiro, 3810-193 Aveiro, Portugal 4
5
Address correspondence to A. Correia, University of Aveiro, Biology Department and 6
CESAM, Campus Universitário Santiago, 3810-193 Aveiro, Portugal. Telephone: 7
(351)234370970. E-mail: [email protected] 8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Copyright © 2012, American Society for Microbiology. All Rights Reserved.Appl. Environ. Microbiol. doi:10.1128/AEM.00359-12 AEM Accepts, published online ahead of print on 6 April 2012
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
2
ABSTRACT 24
We compared the resistome within polluted and unpolluted rivers, focusing on 25
extended-spectrum beta-lactamases (ESBL) genes, in particular blaCTX-M. Twelve 26
rivers from a Portuguese hydrographic basin were sampled. Physicochemical and 27
microbiological parameters of water quality were determined and results classified 9 28
rivers as unpolluted (UP) and 3 as polluted (P). Of the 225 cefotaxime-resistant strains 29
isolated, 39 were identified as ESBL producers, with 18 carrying a blaCTX-M gene (15 30
from P and 3 from UP). Analysis of CTX-M nucleotide sequences showed that 17 31
isolates produced CTX-M from group 1 (CTX-M-1, -3, -15 and -32) and 1 gene 32
belonged to group 9 (CTX-M-14). The genetic environment study revealed the presence 33
of different genetic elements previously described in clinical strains. ISEcp1 was found 34
in the upstream region of all isolates examined. Culture-independent blaCTX-M-like 35
libraries comprised 16 CTX-M gene variants, 14 types in the P library and 4 types in UP 36
library, varying from 68% to 99% similarity between them. Besides the much lower 37
diversity among UP CTX-M-like genes, the majority were similar to chromosomal 38
ESBLs such as blaRAHN-1. The results demonstrate that occurrence and diversity of 39
blaCTX-M genes are clearly different between polluted and unpolluted lotic 40
ecosystems; these findings favor the hypothesis that natural environments are reservoirs 41
of resistant bacteria and resistance genes, where anthropogenic-driven selective 42
pressures may be contributing to the persistence and dissemination of genes usually 43
relevant in clinical environments. 44
45
Keywords: lotic ecosystem, antibiotic resistance, pollution, beta-lactamases, cefotaxime 46
47
48
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
3
Introduction 49
Antibiotics are widely used not only to treat human and animal infections but also in 50
farms and aquacultures as food additives to promote animal growth and prevent 51
diseases. Consequently, antibiotics are released in large amounts in natural ecosystems 52
where they can impact the structure and activity of environmental microbial populations 53
(22, 23). Undoubtedly, the occurrence and dissemination of antibiotic resistant bacteria 54
(ARB) and antibiotic resistance genes (ARGs) are recognized worldwide as a major 55
public health concern. Efforts on prevention of ARGs and ARB spread focused on a 56
clinical and human-community level, being especially centered on infection control and 57
restriction of antibiotic use (34). However, considering the growing evidences that 58
ARGs and pathogenic ARB are no longer restricted to clinical settings, it is quite clear 59
that the research activities need to be expanded to include non-pathogenic 60
environmental microorganisms that could be the potential source for these ARGs (22, 61
23, 36, 38). 62
Aquatic systems can be highly impacted by human activities receiving contaminants 63
and bacteria from different sources and thus encouraging the promiscuous exchange and 64
mixture of genes and genetic platforms. Consequently these systems may promote the 65
spread of ARB and ARGs and even the emergence of novel resistance mechanisms and 66
pathogens (2, 3, 38). Considering the frequent detection of ARGs and ARB in aquatic 67
systems and since their dissemination constitutes a serious public health problem, it has 68
been suggested that ARGs should be considered as environmental emerging 69
contaminants (23, 29). 70
Beta-lactam antibiotics are the most broadly used antibacterial agents. Extended-71
spectrum beta-lactamases (ESBLs) mediate resistance to broad-spectrum beta-lactams 72
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
4
such as cefotaxime and ceftazidime, and are widely disseminated among Gram-negative 73
bacteria. Since first reported in 1983 (19), the occurrence of infections caused by ESBL-74
producing bacteria has been constantly rising and constitutes a serious threat to human 75
health. CTX-M genes have rapidly become the most common ESBL genes mainly 76
because of the genetic platforms responsible for their mobilization and dissemination 77
(insertion sequences, integrons, transposons, plasmids). Particularly common on the 78
genomic environment of these genes are insertion sequences such as ISEcp1, IS26 and 79
ISCR1 (4, 5, 8). CTX-M-15 and CTX-M-14 are the most prevalent enzymes, over 110 80
CTX-M-like ESBLs described so far, mostly found in Enterobacteriaceae but also, for 81
example, in Aeromonas spp., Pseudomonas spp. and Acinetobacter spp. (6, 8, 25, 35). 82
Interestingly, the CTX-M-like ESBLs are thought to have evolved from chromosomal 83
genes of the non-clinical genus Kluyvera (28). Few studies addressed the links between 84
pollution and the dispersal of ARB and ARGs in natural environments. It is of major 85
importance to understand how anthropogenic activities are modulating the resistance 86
gene pool in order to anticipate future impacts and consequences for the environment 87
and public health. Also, ARGs, and specifically those most frequently found in 88
association with pathogenic bacteria such as CTX-M genes, may be key indicators of 89
water quality and may be used to trace the dissemination of multiresistance in aquatic 90
environments. 91
In this study our goal was to compare the cefotaxime resistome within polluted and 92
unpolluted lotic (flowing waters) ecosystems. Specific goals were: 1) to compare the 93
occurrence and phylogenetic diversity of cefotaxime-resistant bacteria and ESBL 94
producers; 2) to detect and characterize the ESBL genes responsible for the resistance 95
phenotype; 3) to compare the diversity of CTX-M-like genes using culture-dependent 96
and culture-independent approaches. 97
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
5
98
Material and methods 99
Sample collection and water quality assessment 100
Water samples were collected in 12 sites from 11 rivers integrated in the Vouga River 101
basin, located in central Portugal (Fig. 1). Table S1 in the supplemental material 102
indicates the Global Positioning System (GPS) coordinates of all sampling locations. 103
Throughout the basin, these water bodies are exposed to different anthropogenic 104
impacts from agricultural, industrial and domestic origins, which results in different 105
levels of superficial water quality from unpolluted to extremely polluted sites (11). 106
Sampling sites were selected in order to include from putative unpolluted to extremely 107
polluted sites. Water was collected in sterile bottles (7L) from 50 cm below the water 108
surface and kept on ice for transportation. To infer as to the water quality, physical, 109
chemical and microbiological parameters were determined according to Portuguese 110
laws (10), which included pH, color, smell, dissolved oxygen, conductivity, 111
temperature, nitrates, chlorides, phosphates, ammonium, chemical oxygen demand, 112
biological oxygen demand, total and fecal coliforms and fecal streptococci. Surface 113
water quality classification was assigned according to regulations given by the national 114
institute of water (www.inag.pt) which sorts water quality in 5 categories from 115
unpolluted to extremely polluted water in accordance with parameters established by the 116
Portuguese law. 117
118
Enumeration and selection of cefotaxime resistant bacteria 119
Water samples were filtered in 0.45-μm-pore-size cellulose ester filters (Pall Life 120
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
6
Sciences, MI, USA), and the membranes placed on MacConkey agar plates 121
supplemented with 8 μg/ml of cefotaxime to select cefotaxime resistant isolates. Also, 122
to determine the proportion of cefotaxime resistant bacteria among the total bacterial 123
population, plates with no antibiotic supplement were used. Plates were then incubated 124
at 37ºC for 16 h. Colony counting was done in triplicate. Individual cefotaxime-resistant 125
colonies were purified and stored in 20% glycerol at −80ºC. 126
127
Molecular typing and identification of cefotaxime resistant isolates 128
Genomic DNA was isolated as previously described (17). BOX-PCR was used to type 129
all isolates as previously described (32). PCR products were loaded in 1.5% agarose 130
gels for electrophoresis. The banding patterns were analyzed with the software 131
GelCompar (Applied Maths, Belgium). Similarity matrices were calculated with the 132
Dice coefficient. Cluster analysis of similarity matrices was performed by the 133
unweighted pair group method using arithmetic averages (UPGMA).Isolates displaying 134
different BOX profiles were identified by 16S rRNA gene sequencing analysis with 135
primers and PCR conditions as previously described (17). PCR products were purified 136
with the JETQUICK PCR purification spin kit (GENOMED, Löhne, Germany) and 137
used as template in the sequencing reactions. Online similarity searches were performed 138
with the BLAST software at the National Center of Biotechnology Information website. 139
140
Antibiotic susceptibility testing and ESBL detection 141
Antimicrobial resistance patterns were determined by the agar disc diffusion method on 142
Mueller–Hinton agar, against 16 antibiotics from 6 classes: beta-lactams (penicillins, 143
monobactams, carbapenems and 1st, 3rd and 4th generation cephalosporins), quinolones, 144
aminoglycosides, phenicols, tetracyclines and the combination 145
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
7
sulfamethoxazole/trimethoprim. Discs containing the following antibacterial agents 146
were used: amoxicillin (10 μg), amoxicillin/clavulanic acid (20 μg/10 μg), ampicillin 147
(10 μg), aztreonam (30 μg), cefepime (30 μg), cefotaxime (30 μg), ceftazidime (30 μg), 148
cephalothin (30 μg), ciprofloxacin (5 μg), chloramphenicol (30 μg), gentamicin (10 μg), 149
imipenem (10 μg), kanamycin (30 μg), nalidixic acid (30 μg), 150
sulfamethoxazole/trimethoprim (25 μg) and tetracycline (30 μg) (Oxoid, Basingstoke, 151
UK). After 24 h of incubation at 37ºC, organisms were classified as sensitive, 152
intermediate, or resistant according to the Clinical Laboratory Standards Institute 153
guidelines (7). Detection of ESBL production was carried out by the double-disc 154
synergy test (DDST) (18) and a clavulanic acid combination disc method, based on 155
comparing the inhibition zones of cefpodoxime (10 μg) and cefpodoxime-plus-156
clavulanate (10/1 μg) discs (Oxoid, UK). Statistical analysis was performed by two-157
sample t-test with a critical P-value set at 0.05. 158
159
ESBLs and integrase genes screening 160
PCR screening was performed for ESBLs genes encoding SHV, TEM, OXA, CTX-M 161
(group 1, 2, 8/25 and 9), GES, VEB and PER, with primer sets and PCR conditions as 162
described elsewhere (9, 16). Integrase screening was performed for intI1, intI2 and intI3 163
genes (9, 16, 24). Genomic DNA of positive control strains was used (16, 24). Each 164
experiment included as negative control a PCR reaction containing water instead of 165
DNA. Amplicons were analyzed by electrophoresis on a 1.5% agarose gel and stained 166
with ethidium bromide. 167
168
Diversity and genetic environment of blaCTX-M genes 169
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
8
Sequencing was done for the blaCTX-M gene fragments amplified from the bacterial 170
isolates. The presence of ISEcp1, IS26, IS5, orf477, IS903 and orf503 in the genetic 171
environment of blaCTX-M was searched by PCR (12,13,31). 172
173
Construction of blaCTX-M gene libraries 174
To investigate further the diversity of the blaCTX-M genes in both polluted and 175
unpolluted environments environmental DNA from water samples was isolated as 176
previously described (17). DNA isolated from all polluted sites was mixed, as also from 177
unpolluted samples. Hence, two clone libraries of blaCTX-M were constructed using 178
the TA Cloning Kit, according to the manufacturer’s instructions (Invitrogen, Carlsbad, 179
CA, USA). The blaCTX-M gene was amplified using the CTX-F and CTX-R primer set 180
(21). Clones were screened by PCR for the presence of fragments with the expected size 181
by using primers targeting the vector. PCR products were purified and sequenced. 182
Similarity searches were performed using BLAST. A phylogenetic tree was obtained 183
using MEGA version 5 (33). The Shannon–Weaver index of diversity (H) was 184
calculated for each library using the formula H = - Σ(ni/N) log(ni/N), where ni is the 185
abundance of each blaCTX type and N is the sum of analyzed clones in each library. 186
187
Nucleotide sequence accession numbers 188
All blaCTX-M genes nucleotide sequences reported in this work have been deposited in 189
the GenBank database under the accession numbers JQ397652–JQ397669 (bacterial 190
strains) and JQ397670–JQ397721 (clone libraries). Also 16S rRNA gene sequences are 191
available with the accession numbers JQ781502-JQ781652. 192
193
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
9
Results 194
Water quality and occurrence of cefotaxime-resistant bacteria 195
From the analysis of all physical, chemical and microbiological parameters (see Table 196
S2 in the supplemental material) and according to Portuguese law (D.L. 236/98) and the 197
surface water quality classification given by the national water institute, from the 12 198
sites under study, 3 sites were classified as polluted (P) and 9 as unpolluted (UP). All 199
three rivers classified as polluted presented a mixed type of pollution, mainly related to 200
exceptionally high values of phosphates and total coliforms (Table S2 in the 201
supplemental material; D.L. 236/98). 202
The total bacterial counts on MacConkey agar in polluted sites was on average 1.9 X 203
105 CFU/100mL of riverine water of which 8.8% grew on MacConkey agar 204
supplemented with cefotaxime (1.7X104 CFU/100mL), and in pristine rivers was on 205
average 0.68 X 105 CFU/100mL of which 0.6% grew on MacConkey agar 206
supplemented with cefotaxime (4.4 X102 CFU/100mL). 207
208
Molecular typing and identification of bacterial isolates 209
Clonal relationships among cefotaxime resistant isolates (n=225) were assessed by 210
BOX-PCR, and 151 isolates displaying unique BOX profiles were selected for further 211
analysis (see Fig. S1 in the supplemental material). Among strains isolated from 212
polluted waters (n= 60), 41.7% were identified as Pseudomonas spp. (P. fluorescens, P. 213
nitroreducens, P. plecoglossicida and P. putida), 35% affiliated with 214
Enterobacteriaceae members and 21.7% with Aeromonas spp... The Enterobacteriaceae 215
members mostly affiliated with Escherichia coli (25%), followed by Enterobacter spp. 216
(8.33%) and with only an isolate each Alcaligens faecalis and Citrobacter freundii. 217
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
10
As of unpolluted waters isolates (n=91) Pseudomonas spp. (P. fluorescens, P. 218
nitroreducens and P. putida) adds 63.7%, Enterobacteriaceae and Aeromonas spp. (A. 219
media and A. hydrophila) with 8.8% and 1.1% respectively, and Acinetobacter sp. 220
appears as the second most abundant genus in these samples, with 26.4% (all 221
Acinetobacter calcoaceticus). Among Enterobacteriaceae members, Enterobacter sp. 222
and Escherichia coli were identified (5.5% and 3.3%, respectively). A 16S rRNA gene 223
phylogenetic tree is presented in supplemental material Fig.S2. 224
225
Antimicrobial susceptibility and detection of ESBL producers 226
As expected, since isolates were selected in agar plates supplemented with cefotaxime, 227
higher numbers of antibiotic resistance were registered for beta-lactams (see Fig. S3 in 228
the supplemental material). It was determined that 22.5 % of the isolates from P and UP 229
samples were resistant to all cephalosporins tested and 52.3% resistant to both 230
cefotaxime and ceftazidime. For beta-lactams, higher percentages (although not 231
statistically significant; two-sample t test, P> 0.05) were always observed for isolates 232
from polluted waters. For non-beta-lactam antibiotics higher resistance levels were 233
observed against quinolones (in particular nalidixic acid with 78.1% resistants), 234
sulfamethoxazole-trimethoprim and chloramphenicol (55% and 51%, respectively). In 235
isolates from polluted environments also resistance to tetracycline (36.7%) and to 236
aminoglycosides (31.7%) was frequently detected. Besides imipenem (99.3% 237
susceptible strains), gentamicin was the most effective, with only 3.3% resistance 238
among isolates from UP and 21.7% from P sites. The less effective were the penicillins, 239
the monobactam aztreonam and 1st and 3rd generation cephalosporins. Significant 240
differences were found among isolates from polluted and unpolluted waters in 241
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
11
resistance frequencies towards aminoglycosides, quinolones, tetracycline and the 242
combination sulfamethoxazole/trimethoprim (two-sample t test, P< 0.05). 243
Multiresistance (defined as resistance to 3 or more classes of antibiotics, including beta-244
lactams) was found in 56.6% and 46.0% of the strains isolated from polluted and 245
unpolluted sites, respectively. 246
Of the 151 isolates tested, 39 were positive for ESBL production by both used methods, 247
with 27 isolates from polluted waters (13 Escherichia coli, 8 Aeromonas spp. and 6 248
Pseudomonas spp.) and 12 isolates from unpolluted sites (7 Pseudomonas spp., 2 249
Acinetobacter sp., 2 Escherichia coli and 1 Aeromonas spp.). 250
251
Occurrence and diversity of integrases and ESBLs genes 252
The ESBL-producing isolates were further analyzed by PCR screening for ESBLs and 253
integrase genes. As for ESBL genes the most frequently detected was blaCTX-M 254
(n=18) followed by blaTEM (n=10). In 6 strains it were identified both blaCTX-M and 255
blaTEM. Two blaVEB were identified, both on Aeromonas sp., once in each 256
environment. OXA-1-like genes were detected in 6 strains isolated from polluted sites. 257
No blaGES, blaPER, blaSHV or blaOXA-2 and -10-like were identified with the 258
primer sets used in this study. 259
Integrase genes intI1, intI2 and intI3 were screened by PCR among the 39 ESBL-260
producers. On 22 out of 39 isolates it was detected intI1 (19 P and 3 UP), affiliated with 261
Escherichia coli (11 P and 1 UP), Pseudomonas sp. (2 P and 1 UP) and Aeromonas sp. 262
(6P and 1UP). The intI2 and intI3 genes were not detected. 263
264
Diversity and genetic environment of blaCTX-M genes 265
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
12
Since blaCTX-M was the most frequently detected among the 39 ESBL-producers, 266
blaCTX-M genes were further characterized (Table 1). 267
The CTX-M genes were detected in 18 isolates (15P and 3 UP). The nucleotide 268
sequence of blaCTX-M genes was determined and their genomic environment was 269
inspected by PCR and sequencing. Sequence analysis showed that isolates produced 270
CTX-M from group 1 (CTX-M-1, -3, -15 and -32) and group 9 (CTX-M-14). The CTX-271
M-1 gene was found in 3 isolates (all from polluted water), CTX-M-3 gene in 3 isolates 272
(all from polluted water), CTX-M-15 in 10 isolates (8P and 2UP) and CTX-M-32 was 273
detected in only 1 isolate from unpolluted water. From group 9 it was found CTX-M-14 274
gene in one strain isolated from polluted water. The genetic environment study revealed 275
the presence of 6 different genetic environments with elements previously described in 276
clinical strains. A schematic representation of the different genomic environments found 277
in the 18 isolates is presented in figure 2. ISEcp1 was found in the upstream region of 278
all isolates examined in the present study, but disrupted in 8 isolates by IS26 and in 1 by 279
IS5. The distance between ISEcp1 and the start codon of blaCTX-M genes was as 280
previously described, varying from 32bp to 127bp. All blaCTX-M from group 1 281
presented downstream an Orf477. The only blaCTX-M from cluster 9 detected was 282
blaCTX-M-14 (E6) which presented downstream an IS903-like element. 283
284
Polluted and unpolluted blaCTX-M-like clone libraries 285
To compare the diversity of blaCTX-M genes in polluted and unpolluted environments, 286
two clone libraries of blaCTX-M-like gene fragments were constructed and analyzed. 287
Gene fragments were amplified using as template two environmental DNA pools 288
corresponding each to P and UP samples. A total of 52 clones were obtained and all 289
inserts were sequenced (27 P and 25 UP). Culture-independent blaCTX-M-like libraries 290
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
13
comprised 16 gene variants (A-P), 14 types in the P library (H= 1.04) and 4 types in UP 291
library (H= 0.23), with similarity values varying from 68% to 99% between them and 292
from 97% to 100% with sequences from GenBank database. The majority (n=16) 293
affiliated with nucleotide sequences of blaCTX-M variants from group 1 (CTX-M-1, -294
12. -15, -30, -37, -68 and -97) but also blaCTX-M from group 2 (CTX-M-97) (n=2), 295
group 9 (CTX-M-14) (n=3) and group 25 (CTX-M-78 and -100) (n=2) were identified. 296
Besides the much lower diversity among UP CTX-M-like genes, the majority were 297
similar to chromosomal ESBLs such as blaRAHN-1, blaRAHN-2 and blaFONA-5 (Fig. 298
3). 299
300
Discussion 301
Lotic ecosystems are threatened daily by anthropogenic actions that compromise water 302
quality and, in consequence, its sustainable use. 303
Considering aquatic systems as reactors for diverse biological interactions that have 304
important genetic implications, the study of the aquatic antibiotic resistome (which 305
includes ARGs, pathogenic and non-pathogenic ARBs) is important, as it might indicate 306
the extent of alteration of water ecosystems by anthropogenic activities. Several studies 307
have been reporting the presence of antibiotic resistant bacteria from several aquatic 308
environments but focusing on pathogenic organisms or directly related to an 309
environmental threat as a hospital sewage discharge (1, 3, 36). 310
In this study, two groups of rivers (polluted and unpolluted), which are part of the same 311
Portuguese lotic ecosystem, were inspected for the presence of cefotaxime-resistant 312
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
14
Gram-negative bacteria, in order to understand how human action is modulating the 313
environmental resistome, in particular the cefotaxime-resistome. 314
As expected, high levels of resistance were obtained in this study, among CTXR 315
isolates, against other beta-lactams frequently conferred by the same resistance 316
mechanism (16), with higher occurrence among P strains. ESBL-production was 317
detected in Pseudomonas sp., Acinetobacter sp., Escherichia coli and Aeromonas sp., 318
and was more frequent among isolates from polluted sites. Recently in several 319
environmental studies, members of the same genera have been identified as ESBL-320
producers, enforcing their relevance and importance for resistance monitoring 321
(14,15,27). We investigated the presence of different ESBL genes and found blaCTX-M 322
gene as the most prevalent followed by blaTEM genes. The majority of the isolated 323
CTX-M-producers affiliated with E. coli but also with Aeromonas hydrophila (3 324
blaCTX-M-3) and Pseudomonas sp. (1 blaCTX-M-15). Few studies have reported the 325
presence of blaCTX-M genes in Pseudomonas spp. and Aeromonas spp.. A previous 326
study reported blaCTX-M-27 genes in 2 Aeromonas sp. isolated in river sediment (21). 327
Also Aeromonas spp. producing blaCTX-M-3 and blaCTX-M-15 have been detected in 328
clinical settings and directly implicated in human infections (37). As far as we know, 329
this is the first work reporting environmental Aeromonas spp. producing blaCTX-M-3 330
genes. Also in Pseudomonas spp. reports on CTX-M producers are rare. In fact the 331
majority refer to clinical Pseudomonas aeruginosa isolates which have been reported to 332
produce CTX-M-1, -2, -15 and –43 (26). Also recently 2 spinach saprophyte strains 333
identified as P. putida and 1 P. teessidea were referred as CTX-M-15-producers (30). 334
To detect any potential genetic platforms able to mobilize the blaCTX-M genes, we also 335
analyzed the genomic environment of the 18 blaCTX-M genes detected. Different 336
insertion sequence elements were found. Upstream the bla gene in all strains it was 337
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
15
detected an ISEcp1 element. Other IS elements (IS5 and IS26) were found but 338
disrupting the ISEcp1 element. The organization IS26 and end of ISEcp1 has been 339
mostly found in clinical Enterobacteriaceae isolates but it was also described in an E. 340
coli blaCTX-M-1 producer isolated from seagulls fecal droppings (12, 27, 31). On the 341
other hand, the organization IS5 and end of ISEcp1 was found upstream the blaCTX-M-342
32 gene in environmental and clinical E. coli isolates (13, 27). The presence of ISEcp1 343
element upstream blaCTX-M-1, -3, -14 and blaCTX-M-15 has been also reported in 344
clinical isolates (12, 20, 31). Downstream of the bla genes in the CTX-M-1 group, 345
sequence ORF477 was present in all strains. Another insertion sequence, IS903, was 346
found downstream the blaCTX-M-14 from CTX-M group 9, as already described by 347
other authors in clinical Enterobacteriaceae isolates (12, 20, 31). The common 348
phenotype of multiresistance among ESBL-producing isolates is a result of the presence 349
of other genes, normally encoded in the same plasmid carrying ESBL genes. This gene 350
panoply contributes to maintaining ESBL-producing bacterial communities, even with 351
low concentration of beta-lactams (8). As reported in this work, it is of particular 352
concern the fact that 88.9% of the CTX-M-producers are multiresistant (93.3% P and 353
66.6% UP). Among CTX-M-producers isolated from polluted waters, resistance to 354
quinolones, aminoglycosides, tetracyclines and the combination sulfamethoxazole-355
trimethoprim was highly prevalent. Due to their ability to capture and incorporate gene 356
cassettes from the environment, integrons have an important role on the spread of 357
multidrug resistance in Gram-negative bacteria. In this work, class 1 integrons were 358
detected in 56.4% of ESBL producers (48.7% in P and 7.7% in UP sites). 359
Analyzing only the cultivable fraction of Gram-negative bacteria in MacConkey agar 360
plates might underestimate the diversity of blaCTX-M gene variants present in the lotic 361
ecosystem under study. To overcome this methodological aspect, it was applied a 362
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
16
culture-independent approach to further analyze the diversity of blaCTX-M genes in 363
both environments. For that, two clone libraries of blaCTX-M gene fragments amplified 364
from polluted and unpolluted environmental DNA were constructed and analyzed. In P 365
library the variety of CTX-M-like genes was much higher than in UP library. This 366
probably is related to higher anthropogenic selective pressures posed by the release of 367
antibiotics and/or antibiotic resistant bacteria. Also other studies have shown that other 368
contaminants can also contribute to the persistence of antibiotics resistance in the 369
environment, like for example heavy metals and disinfectants (22,23). Within P library 370
similarity with blaCTX-M genes from 4 clusters and also with chromosomal variants 371
referred as ancestors of clusters CTX-M-1 and CTX-M-2 was found. Interestingly, the 372
majority of blaCTX-M-like sequences found in unpolluted DNA were similar to 373
chromosomal class A ESBLs that have been described in Rahnella spp. (blaRAHN-1 374
and blaRAHN-2) and Serratia fonticola (blaFONA-5). In a previous work a blaCTX-M 375
library cloned from urban river sediment DNA presented also high diversity of blaCTX-376
M sequences with 13 variants found (21). Overall, results here presented show clear 377
differences in polluted and unpolluted environments. While in unpolluted rivers we 378
found at most 4 variants with the majority related to ancestor chromosomally located 379
genes, in polluted waters up to 14 variants were found (from 4 out of 5 clusters so far 380
identified in CTX-M enzymes). 381
A shift in the distribution of different ESBLs has recently occurred in European clinical 382
settings, with a dramatic increase of CTX-M enzymes over TEM and SHV variants. 383
More than 110 CTX-M variants have been described so far. Due to the high homology 384
with chromosomal beta-lactamases from different Kluyvera species these are now 385
recognized as CTX-M ancestors, such as KLUA-1 from K. ascorbata and KLUG-1 386
from K. georgiana (5). However the diversity we found in polluted sites cannot be 387
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
17
attributed to the presence of bacteria carrying CTX-M ancestral genes. As in clinics, 388
our results suggest that CTX-M genes dominance is correlated to selective pressures 389
imposed by human activities. 390
These findings sustain our hypothesis that anthropogenic activities might modulate the 391
environmental resistance gene pool and promote antibiotic resistance dissemination. 392
Also, we have shown that ESBL genes are a form of environmental pollution, either 393
resulting from the intake of ARGs or ARB from human activities or from the selection 394
of environmental resistant bacteria by subtherapeutic antibiotic doses released into the 395
environment. In our study, ESBL genes were found in genera not included in routine 396
evaluation of water quality, associated with the genetic platforms needed for their 397
mobilization and transfer. Thus, we suggest that data on the occurrence and diversity of 398
ESBL genes, and specifically CTX-M genes, can be used to assess ecosystems health 399
and antibiotic resistance evolution. Yet, more studies on other geographical locations 400
are needed to validate this application. These genes are also good candidates to be used 401
as pollution indicators. To further confirm this potential, source tracking approaches 402
must be conducted to link the presence of CTX-M genes to specific sources of 403
contamination. 404
405
Conclusions 406
The work here presented showed that occurrence and antimicrobial susceptibility 407
profiles of CTXR bacteria are markedly different between polluted and unpolluted lotic 408
ecosystems; the same happens with occurrence and diversity of clinically relevant 409
ESBL genes. Our results validate the hypothesis that anthropogenic impacts on water 410
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
18
environments are modulators of the resistance gene pool and promote dissemination of 411
antibiotic resistance. 412
In addition, it suggests that blaCTX-M-like genes may constitute indicators of pollution 413
by antibiotics, useful to study antibiotic resistance dispersal in aquatic environments. 414
We also conclude that the dissemination of resistance to broad-range antibiotics such as 415
cefotaxime may be at an earlier stage in pristine environments, providing the 416
opportunity to continuing studying the impact of anthropogenic-driven dissemination 417
and evolution. 418
419
Acknowledgements 420
This work was financed by Fundação para a Ciência e a Tecnologia (FCT) through 421
grants SFRH/BD/43468/2008 (M.T.) and SFRH/BPD/63487/2009 (I.H.).The authors 422
wish to thank Juliana Nina and Susana Araújo for their assistance in sample collection. 423
424
References 425
1. Allen, H. K., J. Donato, H. H. Wang, K. A. Cloud-Hansen, J. Davies, and J. 426
Handelsman. 2010. Call of the wild: antibiotic resistance genes in natural 427
environments. Nat Rev Microbiol 8:251-259. 428
2. Ash, R. J., B. Mauck, and M. Morgan. 2002. Antibiotic resistance of Gram-429
negative bacteria in rivers, United States. Emerg Infect Dis 8:713-716. 430
3. Baquero, F., J. L. Martinez, and R. Cantón. 2008. Antibiotics and antibiotic 431
resistance in water environments. Curr Opin Biotechnol 19:260-265. 432
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
19
4. Bush, K., and J. F. Fisher. 2011. Epidemiological expansion, structural studies, 433
and clinical challenges of new beta-lactamases from Gram-negative bacteria. 434
Annu Rev Microbiol 65:455-478. 435
5. Cantón, R., and T. M. Coque. 2006. The CTX-M beta-lactamase pandemic. 436
Curr Opin Microbiol 9:466-475. 437
6. Chen, H., W. Shu, X. Chang, J. A. Chen, Y. Guo, and Y. Tan. 2010. The 438
profile of antibiotics resistance and integrons of extended-spectrum beta-439
lactamase producing thermotolerant coliforms isolated from the Yangtze River 440
basin in Chongqing. Environ Pollut 158:2459-2464. 441
7. CLSI, CLSI 2010. Performance standard for antimicrobial susceptibility testing 442
- Document Approved Standard M100-S20. CLSI, Wayne, PA, USA. 443
8. Coque, T. M., F. Baquero, and R. Canton. 2008. Increasing prevalence of 444
ESBL-producing Enterobacteriaceae in Europe. Euro Surveill 13. 445
9. Dallenne, C., A. Da Costa, D. Decre, C. Favier, and G. Arlet. 2010. 446
Development of a set of multiplex PCR assays for the detection of genes 447
encoding important beta-lactamases in Enterobacteriaceae. J Antimicrob 448
Chemother 65:490-495. 449
10. DL 236/98. . Portuguese Legislation on Water Quality (Decree-Law 236/98). 450
11. DRA-Centro, Direcção Regional do Ambiente-Centro. 1998. Plano de bacia 451
hidrográfica do Rio Vouga, 1ª fase, Análise e diagnóstico da situação de 452
referência, anexos temáticos. Lisboa Portugal.Available: 453
http://www.arhcentro.pt/website/OLD_Plan._Bacia_Hidrográfica/PGH_-454
_Rio_Vouga.aspx [accessed 7 February 2012]. 455
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
20
12. Eckert, C., V. Gautier, and G. Arlet. 2006. DNA sequence analysis of the 456
genetic environment of various blaCTX-M genes. J Antimicrob Chemother 457
57:14-23. 458
13. Fernandez, A., E. Gil, M. Cartelle, A. Perez, A. Beceiro, S. Mallo, M. M. 459
Tomas, F. J. Perez-Llarena, R. Villanueva, and G. Bou. 2007. Interspecies 460
spread of CTX-M-32 extended-spectrum beta-lactamase and the role of the 461
insertion sequence IS1 in down-regulating blaCTX-M gene expression. J 462
Antimicrob Chemother 59:841-847. 463
14. Girlich, D., L. Poirel, and P. Nordmann. 2011. Diversity of clavulanic acid-464
inhibited extended-spectrum beta-lactamases in Aeromonas spp. from the Seine 465
River, Paris, France. Antimicrob Agents Chemother 55:1256-1261. 466
15. Guenther, S., C. Ewers, and L. H. Wieler. 2011. Extended-spectrum beta-467
lactamases producing E. coli in wildlife, yet another form of environmental 468
pollution? Front Microbiol 2:246. 469
16. Henriques, I., A. Moura, A. Alves, M. J. Saavedra, and A. Correia. 2006. 470
Analysing diversity among beta-lactamase encoding genes in aquatic 471
environments. FEMS Microbiol Ecol 56:418-429. 472
17. Henriques, I. S., A. Almeida, A. Cunha, and A. Correia. 2004. Molecular 473
sequence analysis of prokaryotic diversity in the middle and outer sections of the 474
Portuguese estuary Ria de Aveiro. FEMS Microbiol Ecol 49:269-279. 475
18. Jarlier, V., M. H. Nicolas, G. Fournier, and A. Philippon. 1988. Extended 476
broad-spectrum beta-lactamases conferring transferable resistance to newer beta-477
lactam agents in Enterobacteriaceae: hospital prevalence and susceptibility 478
patterns. Rev Infect Dis 10:867-878. 479
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
21
19. Kliebe, C., B. A. Nies, J. F. Meyer, R. M. Tolxdorff-Neutzling, and B. 480
Wiedemann. 1985. Evolution of plasmid-coded resistance to broad-spectrum 481
cephalosporins. Antimicrob Agents Chemother 28:302-307. 482
20. Lartigue, M. F., L. Poirel, and P. Nordmann. 2004. Diversity of genetic 483
environment of bla(CTX-M) genes. FEMS Microbiol Lett 234:201-207. 484
21. Lu, S. Y., Y. L. Zhang, S. N. Geng, T. Y. Li, Z. M. Ye, D. S. Zhang, F. Zou, 485
and H. W. Zhou. 2010. High diversity of extended-spectrum beta-lactamase-486
producing bacteria in an urban river sediment habitat. Appl Environ Microbiol 487
76:5972-5976. 488
22. Martinez, J. L. 2009. Environmental pollution by antibiotics and by antibiotic 489
resistance determinants. Environ Pollut 157:2893-2902. 490
23. Martinez, J. L. 2009. The role of natural environments in the evolution of 491
resistance traits in pathogenic bacteria. Proc Biol Sci 276:2521-2530. 492
24. Moura, A., C. Pereira, I. Henriques, and A. Correia. 2012. Novel gene 493
cassettes and integrons in antibiotic-resistant bacteria isolated from urban 494
wastewaters. Res Microbiol.163: 92-100. 495
25. Novais, A., I. Comas, F. Baquero, R. Canton, T. M. Coque, A. Moya, F. 496
Gonzalez-Candelas, and J. C. Galan. 2010. Evolutionary trajectories of beta-497
lactamase CTX-M-1 cluster enzymes: predicting antibiotic resistance. PLoS 498
Pathog 6:e1000735. 499
26. Picao, R. C., L. Poirel, A. C. Gales, and P. Nordmann. 2009. Further 500
identification of CTX-M-2 extended-spectrum beta-lactamase in Pseudomonas 501
aeruginosa. Antimicrob Agents Chemother 53:2225-2226. 502
27. Poeta, P., H. Radhouani, G. Igrejas, A. Goncalves, C. Carvalho, J. 503
Rodrigues, L. Vinue, S. Somalo, and C. Torres. 2008. Seagulls of the 504
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
22
Berlengas natural reserve of Portugal as carriers of fecal Escherichia coli 505
harboring CTX-M and TEM extended-spectrum beta-lactamases. Appl Environ 506
Microbiol 74:7439-7441. 507
28. Poirel, L., P. Kampfer, and P. Nordmann. 2002. Chromosome-encoded 508
Ambler class A beta-lactamase of Kluyvera georgiana, a probable progenitor of 509
a subgroup of CTX-M extended-spectrum beta-lactamases. Antimicrob Agents 510
Chemother 46:4038-4040. 511
29. Pruden, A., R. Pei, H. Storteboom, and K. H. Carlson. 2006. Antibiotic 512
resistance genes as emerging contaminants:� studies in northern Colorado†. 513
Environ Sci Technol 40:7445-7450. 514
30. Raphael, E., L. K. Wong, and L. W. Riley. 2011. Extended-spectrum beta-515
lactamase gene sequences in Gram-negative saprophytes on retail organic and 516
nonorganic spinach. Appl Environ Microbiol 77:1601-1607. 517
31. Saladin, M., V. T. Cao, T. Lambert, J. L. Donay, J. L. Herrmann, Z. Ould-518
Hocine, C. Verdet, F. Delisle, A. Philippon, and G. Arlet. 2002. Diversity of 519
CTX-M beta-lactamases and their promoter regions from Enterobacteriaceae 520
isolated in three Parisian hospitals. FEMS Microbiol Lett 209:161-168. 521
32. Tacao, M., A. Alves, M. J. Saavedra, and A. Correia. 2005. BOX-PCR is an 522
adequate tool for typing Aeromonas spp. Antonie Van Leeuwenhoek 88:173-523
179. 524
33. Tamura, K., D. Peterson, N. Peterson, G. Stecher, M. Nei, and S. Kumar. 525
2011. MEGA5: molecular evolutionary genetics analysis using maximum 526
likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol 527
Evol 28:2731-2739. 528
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
23
34. Taylor, N. G., D. W. Verner-Jeffreys, and C. Baker-Austin. 2011. Aquatic 529
systems: maintaining, mixing and mobilising antimicrobial resistance? Trends 530
Ecol Evol 26:278-284. 531
35. Woodford, N., J. F. Turton, and D. M. Livermore. 2011. Multiresistant 532
Gram-negative bacteria: the role of high-risk clones in the dissemination of 533
antibiotic resistance. FEMS Microbiol Rev 35:736-755. 534
36. Wright, G. D. 2010. Antibiotic resistance in the environment: a link to the 535
clinic? Curr Opin Microbiol 13:589-594. 536
37. Ye, Y., X. H. Xu, and J. B. Li. 2010. Emergence of CTX-M-3, TEM-1 and a 537
new plasmid-mediated MOX-4 AmpC in a multiresistant Aeromonas caviae 538
isolate from a patient with pneumonia. J Med Microbiol 59:843-847. 539
38. Zhang, X. X., T. Zhang, and H. H. Fang. 2009. Antibiotic resistance genes in 540
water environment. Appl Microbiol Biotechnol 82:397-414. 541
542
543
544
545
546
547
548
549
550
551
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
24
Tables 552
Table 1: Characteristics of the blaCTX-M producers isolated from polluted (P) and 553
unpolluted (UP) samples, regarding phylogenetic affiliation, sample origin, ESBL and 554
integrase genes detected and antimicrobial resistance profile. 555
Isolate
Phylogenetic
affiliation
Sample (P/UP)
ESBL genes
detected by PCR
Antibiotic resistance profile
IntI 1
E1 A. hidrophila. P blaTEM, blaCTX-M AML, AMP, AMC, KF, CTX, FEP, CIP, NA, CN, K, TE +
E2 A. hydrophila P blaTEM, blaCTX-M AML, AMP, AMC, KF, CTX, FEP, NA, CN, K, TE +
E3 A. hydrophila P blaTEM, blaCTX-M AML, AMP, AMC, ATM, KF, CTX, FEP, CIP, NA, K, TE +
E4 E. coli P blaCTX-M AML, AMP, AMC, ATM, KF, CTX, CAZ, FEP, CIP, NA, C, CN, K, TE -
E5 E. coli P blaCTX-M, blaOXA AML, AMP, AMC, ATM, KF, CTX, CAZ, FEP, CIP, NA, CN, K, SXT, TE +
E6 E. coli P blaTEM, blaCTX-M AML, AMP, AMC, ATM, KF, CTX, NA, C, TE +
E7 E. coli P blaTEM, blaCTX-M, blaOXA AML, AMP, AMC, ATM, KF, CTX, CAZ, FEP, CIP, NA, CN, K, SXT, TE +
E8 E. coli P blaCTX-M, blaOXA AML, AMP, AMC, ATM, KF, CTX, CAZ, FEP, CIP, NA, CN, K, SXT, TE +
E9 E. coli P blaCTX-M AML, AMP, ATM, KF, CTX, CAZ, FEP, CIP, NA, CN, K, SXT, TE +
E10 E. coli P blaCTX-M AML, AMP, ATM, KF, CTX, CAZ, FEP, CIP, NA, K, SXT, TE +
E11 E. coli P blaCTX-M AML, AMP, ATM, KF, CTX, CAZ, FEP, CIP, NA, K, SXT, TE +
E12 E. coli P blaCTX-M, blaOXA AML, AMP, AMC, ATM, KF, CTX, CAZ, FEP, CIP, NA, CN, K, SXT, TE +
E13 E. coli P blaTEM, blaCTX-M AML, AMP, AMC, ATM, KF, CTX, CAZ, FEP, SXT +
E14 E. coli P blaCTX-M AML, AMP, ATM, KF, CTX, FEP, SXT, TE +
E15 E. coli P blaCTX-M AML, AMP, AMC, ATM, KF, CTX, CAZ, FEP, NA, SXT, TE +
E16 E. coli UP blaCTX-M AML, AMP, ATM, KF, CTX, CAZ, FEP, CIP, NA, SXT +
E17 E. coli UP blaTEM, blaCTX-M AML, AMP, AMC, ATM, KF, CTX, CAZ, TE -
E18 Pseudomonas sp. UP blaCTX-M AML, AMP, AMC, ATM, KF, CTX, NA, C, SXT +
556
557
558
559
560
561
562
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from
25
Figure legends 563
FIG. 1: Map of Vouga River basin (Central Portugal) with the location of the 12 564
sampling sites under study (1,2 and 12- polluted; 3-9 - unpolluted). 565
566
567
FIG. 2: Schematic representation of the genetic environment of CTX-M genes from the 568
18 isolates producing CTX-M from group 1 (CTX-M-1, -3, -15 and -32) and group 9 569
(CTX-M-14). The number of isolates from each polluted and unpolluted environment 570
that carry each variant is indicated. 571
572
FIG. 3: Dendrogram tree of blaCTX-M gene sequences types A to N identified from the 573
polluted (P) and unpolluted (UP) genomic libraries. The number in parentheses shows 574
the number of times the sequence was found in the library. The branch numbers refer to 575
the percent confidence as estimated by a bootstrap analysis with 1000 replications. 576
577
578
579
580
581
582
583
on March 21, 2021 by guest
http://aem.asm
.org/D
ownloaded from