World J Microbiol Biotechnol (2013) 29:1317–1325 DOI 10.1007/s11274-013-1295-3
ORIGINAL PAPER
Pseudomonas putida are environmental reservoirs of antimicrobial resistance to b-lactamic antibiotics Catarina Meireles • Gonc¸alo Costa • Ineˆs Guinote • Teresa Albuquerque • Ana Botelho • Carlos Cordeiro • Patrick Freire
Received: 15 October 2012 / Accepted: 14 February 2013 / Published online: 19 February 2013 Ó Springer Science+Business Media Dordrecht 2013
Abstract The adaptive flexibility of bacteria largely contributes to the emergence of antibiotic resistance, eventually leading to the predictable failure of current antimicrobial therapies. It is of utmost importance to improve current approaches and implement new ways to control bacterial growth and proliferation. A promising strategy lies in unraveling the antimicrobial resistance (AMR) dynamics in environmental reservoirs, namely in soil. Environmental microorganisms are antibiotic producers and generally also carriers of AMR mechanisms. Therefore, soil samples were collected from areas distinctly influenced by men: rural farms and urban fluvial shores. Globally, microbial communities collected in farms revealed the highest antibiotic resistance potential. Largely predominant Gram-negative isolates were further screened for their low susceptibility to b-lactamic agents, and found to belong to Pseudomonaceae family, with predominance of Pseudomonas putida (92 %). Minimal Inhibitory Concentration (MIC) was determined for five b-lactams and the distributive analysis of cefotaxime MIC performed, allowing the first report of Epidemiological Cut-OFF values for P. putida regarding such antibiotic. Hence, 46 % of the isolates from farms presented acquired resistance to cefotaxime, with fluvial strains presenting an acquisition of AMR in 22 % of the isolates. The response to b-lactams
C. Meireles I. Guinote T. Albuquerque A. Botelho P. Freire (&) Bacteriology Laboratory, Microbiology and Animal Pathology Unit, Instituto Nacional de Investigac¸a˜o Agra´ria e Veterina´ria, Estrada de Benfica 701, 1549-011 Lisbon, Portugal e-mail:
[email protected] G. Costa C. Cordeiro Faculdade de Cieˆncias, Centro de Quı´mica e Bioquı´mica, Universidade de Lisboa, Lisbon, Portugal
impact in P. putida is different from Pseudomonas aeruginosa’s, the family type strain, showing that data determined for a species should only be extended to other bacteria with caution, even closely related. It becomes crucial to broaden present research, mainly focused on few pathogenic bacteria, to other microorganisms carrying relevant resistance tools or capable of genetic transfer to more virulent strains. Most available data on AMR so far has been obtained from studies performed in restricted clinical or veterinary context, showing the result of a strong selective pressure related to therapy but often disregarding the origin of the AMR mechanisms encountered. The strong impact that environmental microorganisms have (and probably already had in the past) on the evolution and spreading of AMR, is just beginning to be unveiled. Keywords Environmental isolates Pseudomonas putida Acquired antibiotic resistance Environmental resistance reservoirs
Introduction The dissemination of antibiotic resistance determinants in pathogenic bacteria is arising as the most relevant current medical setback in the treatment of infectious diseases. The acquisition of the mechanisms underlying antimicrobial resistance (AMR) in bacteria is due in part to the triggering of spontaneous mutations (Martinez et al. 2000), highly selected under pressure of coexistence with antibiotics, but also relates to described cases of horizontal transfer of genetic material between discrete organisms (Davies 1994). Anthropogenic abuse of antibiotics, in clinical settings or also often represented by environmental contaminations, increases the selection of resistant microorganisms.
123
1318
New drug resistant pathogens, sometimes carrying resistances against several antimicrobials, are emerging. Besides clinical settings, the presence of bacteria with AMR has been increasingly detected in animal breeding facilities and slaughterhouses, in soils and wastewater (mostly through faecal cross-contamination), in urban sewage waters and agricultural areas contaminated with pesticides. Biocide substances, extensively used in healthcare and consumer products, have also been correlated with the activation of genes involved in AMR (Baucheron et al. 2002). Moreover, the complexities of some AMR mechanisms, based on enzymatic activities, imply long and selective evolutive procedures, suggesting that the molecular mechanisms conferring resistance to bacteria may have more remote origins and be independent from the last century waves of antibiotics massive use. The latter however favoured a positive selection of resistant strains. Many AMR mechanisms can indeed be traced to microorganisms carrying natural resistance determinants, located in environmental reservoirs. For instance, environmental organisms such as the members of Streptomyces genus, are responsible for the production of a substantial set of the antibiotics nowadays used clinically and are, by default, resistant to their action (Clardy and Walsh 2004). These secreted substances are produced in different ecological niches, a priori to ensure growth inhibition of nutrient competing organisms. Such molecules might also fulfil distinct roles in nature, where antibiotics are normally found at significantly lower concentrations than those used in clinical conditions. Namely, they have been suggested to act as signalling molecules in low concentrations (Fajardo and Martı´nez 2008). The natural presence of antibiotics in these niches leads to the maintenance of a complex, still uncharacterized, environmental reservoir of resistance determinants, carried either by antibiotic producers or by other organisms that prevail in the same surroundings (Riesenfeld et al. 2004; D’Costa et al. 2006). Environmental bacteria have been found to globally present low susceptibility or resistance to several antibiotics (even synthetic) and in addition, to be able to use practically any of all major classes of antibiotics as sole carbon source (D’Costa et al. 2006; Dantas et al. 2008). Finally, the increasing spread of antimicrobial resistance, observed in diverse bacterial populations, may result from the clonal selection of organisms tolerant to sub lethal antimicrobial doses, which present greater fitness under those selective pressures. The survival and adaptation capacities that bacteria show when facing the most diverse conditions can explain in part why they constitute a reserve of potential AMR mechanisms, available in distinct reservoirs of the environment, including animals. Highly stressful or toxic conditions, often from human origin, have furthermore been
123
World J Microbiol Biotechnol (2013) 29:1317–1325
recognized to have an impact on AMR selection and spreading (Beaber et al. 2004). It is thus of the utmost importance to assess the factors that can influence this dissemination of AMR in order to consolidate the basis of a solid strategy to solve the emergent problem of resistant bacteria. During the last decades, the majority of the efforts to combat drug (and multidrug) resistant pathogens have focused mainly on controlling Gram-positive bacteria, with specific novel antimicrobial agents developed and new strategies implemented (Souli et al. 2008). However, the emergence of pan-resistant Gram-negative pathogens, essentially challenging all available drugs, is becoming a serious and unattended threat (Livermore 2004). Most are opportunistic pathogens that present such a high metabolic versatility, that it allows their survival in the most diverse conditions, even extreme. This amazing capacity against a panoply of highly stressful or toxic conditions has also been correlated to the dissemination of genes involved in AMR (Beaber et al. 2004). Amongst all known AMR aspects, resistance to b-lactams is considered a priority target of research since its impact is heavy on reducing treatment efficacy against Gram-negative pathogens. Bacterial resistance to b-lactams in Gram-negative bacteria relies heavily on the production of b-lactamases, however distinct molecular pathways have been determined (Livermore 1995). In this work, the main objective was to screen, identify and characterize resistant environmental strains carrying AMR issued from two different surrounding typologies, exposed to distinct anthropogenic pressures, such as farm environments and urban fluvial shores. The regular and direct use of antimicrobial substances in animal farms for treatment, infections prevention or disinfection, leads to a strong selective pressure on resistant bacteria, specifically derived from antibiotics use. In contrast, urban river shores are relatively free from such a specific pressure; however, through the presence of industrial pollutants and other toxic substances leaked to rivers, constantly renewed and replaced, are subject to strong but variable pressures, leading to the selection of resistant and generally more adaptable strains. The environmental samples for this study were obtained from mud and soils, sporadically contaminated with animal faeces. In this study, the global antibiotic resistance of bacterial communities was assessed and compared between soils from rural farms and urban fluvial shores to deliver a snapshot of the community potential to resist antibiotics as a whole. Gram-negative bacteria with strong levels of tolerance/resistance to b-lactams were then isolated and identified, and their levels of AMR determined by measuring MICs. In order to assess the potential of AMR acquisition in distinct environment locations, the ecological coefficient (ECOFF) for cefotaxime was determined to
World J Microbiol Biotechnol (2013) 29:1317–1325
1319
allow a distinction between acquired and intrinsic resistance in P. putida species, the predominant microorganism selected through the isolation procedure.
visible differences in size or morphology were selected per plate. Each selected colony was subjected to successive plate passage (up to 3 times) to obtain pure cultures.
Materials and methods
Fast assessment of the resistance potential of the environmental communities
Environmental sampling and bacterial isolation Soil and/or mud samples were collected on 11 distinct spots corresponding to two major types, either in urban fluvial shores (type A) or in traditional farms in rural areas (type B). Type A samples were collected in the Tagus river in Lisbon area. Type B samples were collected in areas strongly contaminated with animal faeces and/or feed. They consisted in soil/mud mixed with faeces, taken at traditional farms around Vizela (Fig. 1). The collected material (10 g) was homogenized with sterile water (50 mL). After gravitational settling of the mixture, the liquid fraction was recovered. Material suspensions (including microorganisms) were then collected by centrifugation (12000 g, 5 min). The resulting pellet was resuspended in sterilized water. Primary growth was performed in LB (Luria–Bertani) medium. These cultures were diluted (10-2–10-9) and plated in either LA (LB ? Agar) or LA with ampicillin (8 lg/mL), amoxicillin (8 lg/mL) or cefotaxime (2 lg/mL), to select resistant or reduced susceptibility presenting Gram-negative strains. Ten colonies with
Previous to isolation, the raw liquid fraction suspensions (obtained after processing the soil samples), were inoculated in LB medium (1:1) in 96 wells microplates. Serial 10X dilutions (10-1–10-9) of the original extract were incubated for 16 h at 30 °C. All samples were incubated separately with five antibiotics—ampicillin (8 lg/mL), cefotaxime (2 lg/mL), ciprofloxacin (1 lg/mL), gentamycin (4 lg/mL) and tetracycline (2 lg/mL)—or only growth medium as control, according to the description above. Growth was determined spectrophotometrically at 600 nm. The dilution corresponding to absence of bacterial growth represented the dilution/condition in which no bacteria from the collected community are able to grow, either by dilution (control) or by the action of the antibiotic used. The ratio of the difference between growth without antibiotic (control) and growth in the presence of antibiotic screening indicates the potential of the collected community to resist to the antibiotic tested. A ratio value of 1 thus indicates a bacterial community resistance potential similar with and without antibiotic, indicating strong tolerance to a given antibiotic of bacteria present in the tested community. Identification of isolated strains Detection of Gram-negative strains was performed in selective McConkey nr3 medium with cicloheximide. Biochemical characterization was performed to establish broad strain characteristics. TSI (Triple Sugar Iron), oxidase and catalase tests inferred the ability to ferment dextrose, lactose, sacarose, and sulphured compounds and to produce oxidases and/or catalases (Hajna 1945). According to the previous determinations, a commercially available phenotypic identification system was used to perform the accurate identification of the Gram-negative bacteria isolated. The APIÒ (API 20E and API ID32 GN) test strips were used, coupled with an automated system and software (BioMe´rieux), providing identification with a precision C99.5 %. To select Gram-negative strains, growth was performed on medium MacCONKEY AGAR Nr. 3 (Oxoid).
Fig. 1 Geographical locations of sampling areas, in Portugal. Samples from type A were collected in the Tagus estuary shores, around Lisbon, while type B samples were collected from traditional farms soil around Vizela, northern Portugal
MIC and ECOFF determination Minimum Inhibitory Concentration (MICs) was determined by agar dilution and disk diffusion according to
123
1320
CLSI standards (CLSI, M02-A10; CLSI, M100-S21). Briefly, LA plates were prepared supplemented with serial dilutions of the antibiotic to be tested to determine the MIC. The lowest antibiotic concentration preventing growth of the strain is considered the MIC value. The strain Escherichia coli ATCC 25922 was used as control strain, as recommended. ECOFF value for an antibiotic was determined, according to EUCAST directives (Kahlmeter et al. 2003), by analysis of the distribution of MICs of the strains in a group of strains. It represents the concentration of antibiotic that defines the limit between a ‘‘wild-type’’ strain and a strain with acquired resistance.
Results Sampling was performed in distinct locations to avoid crossed influences and according to two major area types: urban estuarine shores in the Lisbon area—type A—and soils of rural traditional farms around Vizela, northern Portugal—type B—(Fig. 1). Samples consisted in soil and mud, sporadically contaminated with animal faeces, especially in type B locations. The processed raw liquid fraction of the samples collected should be expected to enclose a representative population of microorganisms from the bacterial community at the sampling spot. It will however exclude microorganisms that are strongly aggregated to solid soil particles, and those that cannot grow in laboratory conditions will also be unrecoverable. It has indeed been estimated that only around 1 % of the total microbial population sampled from environmental locations can be isolated in laboratory settings (Wright 2010). Nevertheless, in order to determine a snapshot of general antibiotic resistance profile in the recoverable bacterial communities collected, their global resistance was determined. A simple method was newly established to assess the community global AMR. Determination of the AMR potential of collected microbial communities The resistance potential of the collected bacterial communities was assessed by challenge with antibiotics, to estimate the response of the bacteria from the sample in the presence of a given antibiotic. Such analysis should be representative of the antibiotic resistance potential carried by bacteria in the local cultivable community. Bacterial cultures were serially diluted and the lowest concentration able to grow in the presence of the antibiotics used for this screening was determined (see Materials & Methods). Since the samples are highly heterogeneous, data related to the growth in the presence of antibiotics was normalized by the growth
123
World J Microbiol Biotechnol (2013) 29:1317–1325
Fig. 2 Microbial communities resistance profiles by viability determination. Raw microbial fractions removed from samples were inoculated in LB medium (1:1) in 96 wells microplates. Serial 10X dilutions (10-1–10-9) were incubated for 16 h at 30 °C. All samplings were incubated separately with five antibiotics: Amp ampicillin (8 lg/mL), Cfx cefotaxime (2 lg/mL), Cip ciprofloxacin (1 lg/mL), Gen gentamycin (4 lg/mL) and Tet tetracycline (2 lg/mL), and growth medium as control. Bacterial growth was determined spectrophotometrically at 600 nm. A ratio value of 1 represents 100 % tolerance of the sampled community to that antibiotic (or its concentration). Samples A1–A6: fluvial origin. Samples B1–B5: rural/farm origin
without antibiotics for each sample used as reference. Thus, a general snapshot of the communities AMR potential of viable and cultivable strains was obtained (Fig. 2). Five antibiotics were used, in representation of the most used classes of antibiotics, either in animal and human treatments or prevention. Therefore, the data corresponds to the collective bacterial growth in the presence of ampicillin (8 lg/mL), cefotaxime (2 lg/mL), ciprofloxacin (1 lg/mL), gentamycin (4 lg/mL) and tetracycline (2 lg/mL) determined for the different extracted communities. A ratio value of 1 (growth without antibiotic/growth in the presence of antibiotic), indicates a strong resistance potential among that community. Although these data do not allow quantitative assessment of resistant bacteria, due to the complexity of the bacterial populations, it is clear that samples collected from farms carry overall stronger AMR potential when compared to fluvial urban samples. In type B samples, growth in the presence of tetracyline, ampicillin, gentamycin and ciprofloxacin is similar to the control without antibiotic, indicating high tolerance from bacteria in these samples to these antibiotics, still largely used for animal growthpromotion and treatment. Among urban samples, the sample A4 was exceptional presenting significant resistance potential, essentially to cefotaxime, a third-generation cephalosporin and ampicillin, also a b-lactamic antibiotic but from the first generation. Selection and isolation of environmental Gramnegative, with high tolerance to b-lactamic antibiotics When incubated in a rich non-selective medium, all cultivable bacteria without special requirements would be expected to initiate growth and, in optimal conditions
World J Microbiol Biotechnol (2013) 29:1317–1325
(where no limitations of nutrients and available space exist) a representative community of cultivable microorganisms from the sample location should be attained. However, in a more realistic approach, the colonies obtained represent the population integrating the most adaptable and faster growing cultivable bacteria. Although the faster growing strains always have numeric advantage over slow growing ones, when plated with a dilution cascade, all colonies should be allowed to grow ‘‘independently’’ from the neighbour bacteria, when enough time is allocated. Therefore, by picking a selection of phenotypically different colonies, a representative collection of the faster growing and more versatile sampled isolates was achieved. It was possible to isolate resistant bacteria from all the different sampling sources, even when the global community potential for resistance was estimated as low as seen above. From the 11 sampling spots, 257 isolates, from which all but 3 were Gram-negative bacteria (data not shown), were recovered after screening with either one of three b-lactamic antibiotics [ampicillin (8 lg/mL), amoxicillin (8 lg/mL) and cefotaxime (2 lg/mL)], with concentrations used based on average MIC for Gram-negatives from EUCAST databases. These 254 Gram-negative isolates were identified using APIÒ galleries and their MICs were determined for each of the 3 screening antibiotics (data not shown). Data from antibiograms, strain identification and sampling origin were integrated in order to retain only distinct strains and avoid working with clonal isolates. A final collection of 71 distinct isolates was selected and further characterized in this study. The identifications of these final isolates are represented in Fig. 3. The prevalent genus selected by antibiotic screening was Pseudomonas or closely phylogenetically related genres. Concretely, 92 % of the isolates were identified to the species level as P. putida, corresponding to 65 environmental strains, 35 isolated from fluvial urban shores and 30 from rural area. Furthermore, to estimate the relevance of the antibiotic selection on the screening of resistant strains, other microorganisms were also isolated in parallel from the raw communities without any antibiotic selection. The isolates obtained, broadly representative of the cultivable bacterial community present in the sampling location were identified at least to the genus level, as above. At least nine different genus were thus detected: Pseudomonas, Stenotrophomonas, Streptomyces, Alcaligenes, Aeromonas, Acinetobacter, Bacillus, Escherichia and Rhodococcus.
1321 1% 1% 2%
1% 3%
P. putida P. aeruginosa P. stutzeri S. maltophilia P. fluorescens P. mendocina 92%
Fig. 3 Final isolates identification. Description and composition of the collection of final isolates screened by growth in the presence of screening b-lactamic antibiotics
P. putida isolates was characterized for their resistance to b-lactamic antibiotics. Susceptibility to 6 b-lactams was tested [ampicillin (AMP), amoxicillin (AMX), cefoxitin (FOX), cefpodoxim (CPD), cefpodoxim ? clavulanic acid (CPD ? CLA), cefotaxime (CTX), and ceftazidime (CTZ)] either by agar dilution or by disc-diffusion (CLSI, M02-A10) in all P. putida (see Table 1). A double-disc diffusion test (using CPD in presence or absence of CLA) was also performed to determine variations in the inhibition hallo, in order to detect eventual Extended Spectrum Beta-Lactamase (ESBL) production by the isolates (CLSI, M02-A10). All isolates presented resistance to high concentrations of AMP, AMX and FOX, with MIC values reaching B512 lg/mL for these 3 antibiotics. In addition, all strains were strongly resistant both to CPD and CPD ? CLA, with bacterial growth until the edge of the antibiotic loaded disc. Therefore, no evidence of ESBL production could be inferred from that assay. The resistance of P. putida isolates to CTX is characterized by MICs mostly of 32 lg/mL, however isolates are able to grow at a higher concentration of 64 lg/mL. CTZ has a strong impact on P. putida isolates, only nine isolates presented a MIC of 8 lg/mL, while all other strains showed reduced MIC\4 lg/mL.
Antibiograms of Pseudomonas putida collection
Cefotaxime MIC distribution and estimation of Epidemiological Cut-OFF (ECOFF) of Pseudomonas putida
Due to the large predominance of P. putida among the final isolates, and the reduced information regarding the behavior of this species in AMR context either in clinical, veterinary or environmental areas, a collection of 65 environmental
A major issue in the analysis of antimicrobial resistance data is the determination of resistance breakpoints, corresponding to MIC values that define the distinction of a susceptible strain from a resistant one. The most used in
123
1322
World J Microbiol Biotechnol (2013) 29:1317–1325
Table 1 MIC determination of P. putida isolates Strain #
CTX (lg/mL)
Table 1 continued
AMP (lg/mL)
AMX (lg/mL)
CTZ (lg/mL)
FOX (lg/mL)
Strain #
CTX (lg/mL)
AMP (lg/mL)
AMX (lg/mL)
CTZ (lg/mL)
FOX (lg/mL)
2
32
512
512
\4
\ 512
55
128
256
512
\4
512
3
16
256
512
\4
\ 512
56
64
256
512
8
512
4
32
256
512
\4
512
57
128
256
512
8
512
5
8
512
512
\4
\ 512
58
128
256
512
8
512
6
32
256
512
\4
512
59
64
256
512
8
512
7
32
512
512
\4
512
60
64
256
512
8
512
8 10
32 32
256 256
512 512
\4 \4
512 512
61
128
256
512
8
512
62
64
256
512
\4
512
11
16
256
512
\4
\512
12
32
256
512
\4
512
63 64
128 64
256 256
512 512
\4 \4
512 \512
13
8
256
512
\4
\512
65
32
256
512
\4
512
14
32
256
512
\4
512
66
32
256
512
\4
512
15
32
256
512
\4
512
67
32
256
512
\4
512
16
64
256
512
\4
512
68
32
256
512
\4
512
17
32
256
512
\4
512
69
32
256
512
\4
512
18
64
256
512
\4
512
70
32
256
512
\4
512
19
64
256
512
8
512
71
32
256
512
\4
512
20
64
256
512
4
512
21
64
256
512
\4
512
22
32
256
512
\4
512
23
32
256
512
\4
\512
24 25
32 32
256 256
512 512
\4 \4
512 512
26
32
256
512
\4
\512
27
32
256
512
4
\512
28
32
256
512
\4
512
29
32
256
512
\4
\512
30
16
256
512
\4
512
32
32
256
512
\4
\512
35
32
256
512
\4
512
36
8
256
512
\4
\512
37
32
256
512
\4
512
39
64
256
512
\4
512
40
32
256
512
\4
512
41
32
256
512
\4
512
42
64
256
512
\4
512
44
64
256
512
4
512
45 46
64 32
256 256
512 512
\4 \4
512 512
47
32
256
512
4
512
48
64
256
512
\4
512
49
32
256
512
\4
512
50
32
128
512
8
512
51
32
128
512
\4
512
52
32
256
512
\4
512
53
64
256
512
8
512
54
32
256
512
\4
512
123
CTX Cefotaxime, AMP Ampicilin, AMX Amoxicillin, CTZ Ceftazidime and FOX Cefoxitine resistance data for 65 environmental isolated Pseudomonas putida strains. Data from Cefpodoxime and cefpodoxime ? clavulanic acid are not represented since no inhibition hallo was observed using antibiotics discs
health-related monitorization is the clinical breakpoint, defined by specialized committees, such as EUCAST in Europe (www.eucast.org), which defines if a certain strain is resistant or tolerant to an antibiotic. Information is gathered from all available sources to establish such breakpoints, which can vary yearly according to the available gathered data. Another breakpoint determination, more important from a microbiological point of view, is the ECOFF that corresponds to a MIC value defining the cut-off between a wild type strain and bacteria having acquired AMRs (Kahlmeter et al. 2003). While the clinical breakpoints depend on the regular updating of monitorization of resistance emergence, the ECOFF can be defined for a species and should not vary due to sources, origin or sampling. A strain with a MIC value above the ECOFF is considered as having acquired resistance mechanisms. Even though P. putida is recognized as an opportunistic pathogen of humans and animals, it is still considered a minor health hazard due to the few case reports published. Thus, EU surveillance plans for AMR do not contemplate the collection of data regarding this species. No clinical breakpoint or ECOFF values have been defined by EUCAST for P. putida yet. Nevertheless, our data shows the extreme importance of this bacterium as an environmental reservoir of AMR, highlighting the need to
World J Microbiol Biotechnol (2013) 29:1317–1325
1323
Discussion
Fig. 4 Cefotaxime Minimal Inhibitory Concentration (MIC) distribution per Pseudomonas putida isolate. Distribution of Cefotaxime MIC data for 65 isolated Pseudomonas putida strains to allow determination of ECOFF
further understand its role and potential in the global emergence of AMR. Even though estimation of the clinical breakpoint requires thousands of isolates, it is however possible to estimate the ECOFF values of this species towards the antibiotics tested based on less isolates, since its estimation is independent from the number of isolates analyzed, and thus infer on acquired resistance mechanisms by P. putida among the isolates that constitute this environmental collection. The CTX MIC distribution for 65 isolates of P. putida collected during this study was performed and is presented in Fig. 4. Based on EUCAST criteria (Kahlmeter et al. 2003), the ECOFF value for CTX of P. putida strains can thus be estimated as WT B64 mg/mL.
Farm isolates are more prone to acquire cefotaxime resistance than fluvial isolates The values estimated for CTX ECOFF of P. putida allows the distinction between strains carrying intrinsic AMR mechanisms of that species and strains that have acquired novel mechanisms either by horizontal transfer, capture of external genetic material, or due to the occurrence of mutations in genes already present. Even though data regarding P. putida clinical breakpoints for antimicrobial resistance is still unavailable, we can nevertheless determine the isolates from this environmental collection that present antibiotic resistance features resulting from acquisition of mechanisms of resistance. Based on the data from Table 1, it can thus be estimated that 14 P. putida isolates from farms have acquired AMR for CTX, while only 8 acquired AMR for CTX among fluvial strains. This corresponds to 46 % of farm isolates presenting acquisition of an AMR to CTX, while only half, around 22 % of fluvial strains show that behaviour.
The estuarine ecosystem sampled as type A is the Tagus estuary, located near Lisbon, with a surface area of 320 km2 and a mean volume of 1900 9 106 m3 (Ferreira and Ramos 1989). This estuary plays an important role as an over-wintering area and feeding ground for birds and was partially established as a nature reserve (The Tagus Estuary Nature Reserve). In addition, some areas have been granted a special protection status through the Birds Directive 79/409/EEC. It is thus an area that combines both a strong presence of wild animals, with migratory characteristics, and the pollution inherent of a large river bordered by industrial and agricultural infrastructures, in an urban environment. Furthermore, maritime transportation through the Lisbon harbour and the location of the city in higher grounds strongly impacts this area in terms of pollution. Type B samples were collected in distinct traditional farms around Vizela area, in northern Portugal, much more specialized productive areas and away from major urban areas. These small farms harbour bovines, ovine, caprine and poultry, often bred outdoors and without clear boundaries between animals, leading to a mix of faecal depositions and uncontrolled cross-contaminations between species and soils. However, any pressure exerted is much more sustained and based on the use of growth promoting and prophylactic compounds. The potential AMR contained in the bacterial communities extracted in the first phase of the isolation procedure was assessed. Such preliminary estimation of environmental communities resistance was performed using a novel simple scanning method, based on growth of the bacterial extracted populations, that allows a fast determination of microbial communities potential to tolerate the presence of antibiotics, but could also be used in initial screenings of toxic compounds or variations of growth conditions. Challenge with representatives of several antibiotic classes was performed. The microbial communities collected in farms were shown to contain overall more potential AMR capacity when compared to the ones collected in urban locations. Although such data do not allow quantitative assessment of resistant bacteria, due to the complexity of the bacterial populations, it is clear that samples collected from farms (B) present overall broader AMR potential when compared to fluvial urban samples. In type B samples, growth in the presence of tetracyline, ampicillin, gentamycin and ciprofloxacin was found to be close to the levels of the control batch without antibiotic in different sampling areas, indicating high AMR potential to these antibiotics, largely in use for animal growth-promotion and treatment. However some urban locations presented more specialized resistance profile, probably due to local characteristics of the sampling area, generally
123
1324
determined by local presence or discharge of compounds or industrial leftovers. Among urban samples (A), however, the sample A4 was exceptional presenting significant resistance potential, essentially to cefotaxime, a thirdgeneration cephalosporin and ampicillin, also a b-lactamic antibiotic but from first generation. This sample was collected close to a discharge collector of city sewage, therefore with a strong potential for ‘‘cross-contamination’’ with faeces from several urban animals or even with the urban global sewage system. Other fluvial samples generally presented a broad poor AMR potential. The positive pressure on the selection of antibiotic resistant bacteria by anthropogenic use, especially by the abuse of antibiotics in non-clinical situations such as growth-promotion or prophylaxis in animal production becomes evident when comparing samples from locations differently impacted by antibiotics regular use, leading to the prevalence of bacteria carrying AMR mechanisms directed at these used substances. It was nonetheless possible to isolate resistant bacteria from all sampling sources. From a total of 254 Gram-negative bacterial isolates, a collection of 71 non-clonal isolates was selected and studied; mostly identified as Pseudomonas spp or closely related genres, known for their fast growth characteristics, versatile adaptation to different conditions, and established resistance to stresses, including antibiotics. Among all the mixed populations, Pseudomonas spp showed the strongest potential to be an environmental reservoir of antibiotics resistance mechanisms, either in farms or fluvial shores. Without antibiotic selection, several bacterial genus were isolated from the original samples, clearly illustrating the complex microbial diversity collectable, and showing that the antibiotics used to screen for resistant strains allowed a strong enrichment in bacteria carrying AMR mechanisms. In fact, 92 % of the isolates obtained were identified as P. putida strains. P. putida is ubiquitous in the environment and it can act as an opportunistic pathogen, essentially related to nosocomial infections, presence in hospital surfaces and locations with poor sanitary conditions (Carpenter et al., 2008). P. putida clinical isolates were also found to carry plasmids containing metallo-b-lactamases (MBL) encoding genes, namely blaVIM2, suggesting a genetic transfer of resistance to P. aeruginosa clinical isolates (Juan et al. 2010). Antibiograms were performed for a group of 65 P. putida strains, the prevalent species, in 35 strains isolated from fluvial urban shores and 30 strains from rural area. All 65 strains were tested for the production of ESBL producers but due to the high levels of resistance detected, no determination was possible using CPD and CPD ? CLA. The isolates presented resistance to high concentrations of AMP, AMX and FOX suggesting an intrinsic resistance of P. putida environmental strains to these antibiotics. In addition, all strains were fully resistant both to CPD or CPD ? CLA, with bacterial growth until the edge of the
123
World J Microbiol Biotechnol (2013) 29:1317–1325
antibiotic loaded disc. Regarding cephalosporins, the resistance of P. putida isolates to CTX is characterized by MICs mostly of 32 lg/mL, however a significant number of isolates are able to grow at a concentration 64 lg/mL. Growth at such high concentrations of CTX could be expected for Pseudomonas spp due to the intrinsic resistance to this antibiotic attributed to the closely related P. aeruginosa. Unlike the previous described antibiotics, CTZ has a strong impact on P. putida isolates. Nine isolates presented a MIC of 8 lg/mL, while all other strains have a MIC \4 lg/mL. The MICs determined for AMP, AMX and FOX susceptibilities indicate an intrinsic resistance of P. putida to these 3 antibiotics. Analysis of antimicrobial resistance data is based on the establishment of resistance breakpoints, corresponding to MIC value that distinguishes a susceptible strain from a resistant one. Among breakpoints, the Clinical breakpoint and the Epidemiological Cut-OFF (ECOFF) define resistance or susceptibility of a strain, and acquisition or not of resistance mechanisms, respectively. Based on the data presented in this report, a distribution of the MICs for CTX of the 65 isolates of P. putida collected during this study was performed and is presented in Fig. 4. Based on EUCAST criteria (Kahlmeter et al. 2003), the ECOFF value for CTX of P. putida environmental strains can thus be estimated as WT B64 mg/mL. Even though P. aeruginosa is closely related to P. putida, the latter is intrinsically more resistant to cefotaxime. Data presented on EUCAST database (www. eucast.org) estimates for P. aeruginosa an ECOFF of WT B32 lg/mL for CFX. Therefore, the knowledge generated by the large amount of research studies performed with P. aeruginosa in antimicrobial resistance should only be extended to other Pseudomonaceas with caution. Our data clearly shows significant differences that can have a strong impact on the efficiency of therapeutic strategies in situations like human or animal infections, or on the methods used to maintain proper sanitary conditions in animal production facilities or others. The values estimated for CTX ECOFF of P. putida allow the distinction between strains carrying intrinsic AMR mechanisms of that species and strains that have acquired novel mechanisms either by horizontal transfer, capture of external genetic material, or due to the occurrence of mutations in genes already present. Even though data regarding P. putida clinical breakpoints for antimicrobial resistance is still unavailable, we can nevertheless determine the isolates from this environmental collection that present antibiotic resistance features resulting from acquisition of mechanisms of resistance. 33 % of all isolated P. putida present acquired resistance to CTX, with 46 % of farm isolates presenting acquisition of AMR to CTX, while less than half, around 22 % of fluvial strains show that behaviour. Even though the fluvial strains are
World J Microbiol Biotechnol (2013) 29:1317–1325
subject to a broad range of pressure, from chemical pollution to urban contamination, farm isolates are strongly selected by the constant use of antibiotics related to animal production, and present a very strong potential for the acquisition, maintenance and propagation of specialized AMR strains. This report constitutes, to our knowledge, the first contribution regarding environmental P. putida AMR potential. It is clear that these strains are able of adapting easily to the most diverse situations and antibiotics, as much or even better than P. aeruginosa. Because of those characteristics, there is a strong possibility that this species can constitute a reservoir of transferable resistance mechanisms as well as a strong candidate to an environmental ‘‘factory’’ for the development of new ones. Similar hipothesis has previously been suggested for clinical MBL-producing P. putida isolates, revealing furthermore a possibility of genetic exchange with P. aeruginosa (Juan et al. 2010). Therefore, our data reinforces the need to detect and control not only MBL-producing P. putida in clinical settings, but also to monitor strains carrying other relevant AMRs in the environmental context. It becomes thus important to search for AMR outside the locations were resistant pathogens are mostly found, such as clinical settings or even in the urban community. Such a line of research on AMR must integrate the environmental component, often disregarded, and becomes essential to unveil the origin of bacterial mechanisms and thus implement new antibacterial therapies, before the emergence of new resistance mechanisms. Parallel strategies or even radically new ones are needed to be used in complementarily to classical ones, and re-balance the odds in this endless race to control bacterial proliferation. Acknowledgments Ineˆs Guinote and Gonc¸alo Costa are recipients of post-doctoral support grants by Fundac¸a˜o para a Cieˆncia e Tecnologia (FCT), respectively SFRH/BPD/74711/2010 and SFRH/BPD/ 73779/2010. Patrick Freire is holder of a Researcher Contract by FCT (Cieˆncia 2008 program).
References
1325 fluoroquinolone resistance in Salmonella enterica serovar typhimurium phage type DT204. Microb Drug Resist 8(4):281–289 Beaber JW, Hochhut B, Waldor MK (2004) SOS response promotes horizontal dissemination of antibiotic resistance genes. Nature 427:72–74 Carpenter RJ, Hartzell JD, Forsberg JA, Babel BS, Ganesan A (2008) Pseudomonas putida war wound infection in a US Marine: a case report and review of the literature. J Infect 56(4):234–240 Clardy J, Walsh C (2004) Lessons from natural molecules. Nature 432:829–837 Dantas G, Sommer MO, Oluwasegun RD, Church GM (2008) Bacteria subsisting on antibiotics. Science 320:100–103 Davies J (1994) Inactivation of antibiotics and the dissemination of resistance genes. Science 264:375–382 D’Costa VM, McGrann KM, Hughes DW, Wright GD (2006) Sampling the antibiotic resistome. Science 311:374–377 Fajardo A, Martı´nez JL (2008) Antibiotics as signals that trigger specific bacterial responses. Curr Opin Microbiol 11:161–167 Ferreira JG, Ramos L (1989) A model for the estimation of annual production rates of macrophyte algae. Aquat Bot 33:53–70 Hajna AA (1945) Triple-sugar iron agar medium for the identification of the intestinal group of bacteria. J Bacteriol 49(5):516–517 Juan C, Zamorano L, Mena A, Albertı´ S, Pe´rez JL, Oliver A (2010) Metallo-beta-lactamase-producing Pseudomonas putida as a reservoir of multidrug resistance elements that can be transferred to successful Pseudomonas aeruginosa clones. J Antimicrob Chemother 65(3):474–478 Kahlmeter G, Brown DFJ, Goldstein FW, MacGowan AP, Mouton ¨ Sterlund A, Rodloff A, Steinbakk M, Urbaskova P, JW, O Vatopoulos A (2003) European harmonization of MIC breakpoints for antimicrobial susceptibility testing of bacteria. J Antimicrob Chemother 52:145–148 Livermore DM (1995) beta-Lactamases in laboratory and clinical resistance. Clin Microbiol Rev 8(4):557–584 Livermore DM (2004) The need for new antibiotics. Clin Microbiol Infect 10:1–9 Martinez JL, Baquero F, Mutation (2000) Frequencies and antibiotic resistance. Antimicrob Agents Chemother 44(7):1771–1777 Riesenfeld CS, Goodman RM, Handelsman J (2004) Uncultured soil bacteria are a reservoir of new antibiotic resistance genes. Environ Microbiol 9:981–989 Souli M, Galani I, Giamarellou H (2008) Emergence of extensively drug-resistant and pandrug-resistant gram-negative bacilli in Europe. Euro Surveill 13(47) pii: 19045 Wright GD (2010) Antibiotic resistance in the environment: a link to the clinic? Curr Opin Microbiol 13(5):589–594 CLSI, M02-A10. Performance standards for antimicrobial disk susceptibility tests; approved standard—10th edn CLSI, M100-S21. Performance standards for antimicrobial susceptibility testing; Twenty first informational supplement
Baucheron S, Imberechts H, Chaslus-Dancla E, Cloeckaert A (2002) The AcrB multidrug transporter plays a major role in high-level
123
