Abstract
Raw food is a reservoir of Pseudomonas isolates that could be disseminated to consumers. The presence of Pseudomonas spp. was studied in food samples, and the phenotypic and genotypic characterizations of the recovered isolates were analyzed. Two samples of meat (3%, turkey and beef) and 13 of vegetables (22%, 7 green peppers and 6 tomatoes) contained Pseudomonas spp. A total of 20 isolates were identified, and were classified as follows (number of isolates): P. aeruginosa (5), P. putida (5), P. nitroreducens (4), P. fulva (2), P. mosselli (1), P. mendocina (1), P. monteilii (1), and Pseudomonas sp. (1). These 20 Pseudomonas isolates were clonally different by pulsed-field-gel-electrophoresis, and were resistant to the following antibiotics: ticarcillin (85%), aztreonam (30%), cefepime (10%), imipenem (10%), and meropenem (5%), but were susceptible to ceftazidime, piperacillin, piperacillin-tazobactam, doripenem, gentamicin, tobramycin, amikacin, ciprofloxacin, norfloxacin, and colistin. Only one strain (Ps158) presented a class 1 integron lacking the 3′ conserved segment. The five P. aeruginosa strains were typed by multilocus sequence typing in five different sequence-types (ST17, ST270, ST800, ST1455, and ST1456), and different mutations were detected in protein OprD that were classified in three groups. One strain (Ps159) showed a new insertion sequence (ISPa47) truncating the oprD gene, and conferring resistance to imipenem.
Introduction
V
Pseudomonas is a nonfermenting Gram-negative bacterium that colonizes different terrestrial and aquatic environments (Henry and Speert, 2011), and includes a diversity of species. Among them, Pseudomonas aeruginosa is of great relevance in human medicine, due not only to its frequent implication in nosocomial infections or in chronic lung infections in cystic fibrosis patients, but also to its frequent broad antimicrobial resistance profile (Breidenstein et al., 2011; Henry and Speert, 2011; Olivares et al., 2013). P. aeruginosa has a high intrinsic resistance to several antibiotic families and, on the other hand, an extraordinary capacity to acquire new mechanisms of resistance, even during antibiotic treatment, making it difficult to find alternative therapies (Breidenstein et al., 2011). The intrinsic resistance in P. aeruginosa is due to its low outer membrane permeability, its chromosomal and inducible AmpC β-lactamase, and its complex system of antimicrobial efflux pumps. In addition, P. aeruginosa has the ability to develop resistance by horizontal acquisition of resistance genes carried on plasmids, transposons, or integrons (Rodríguez-Martínez et al., 2009; Breidenstein et al., 2011; Henry and Speert, 2011).
While P. aeruginosa is an important opportunist pathogen responsible for worldwide nosocomial infections, other species such as P. syringae is wide known as being deleterious to plants, and P. putida or P. fluorescens are considered plant growth promoters (Mehri et al., 2011).
Previous studies have given information about the detection of Pseudomonas spp. in food of animal or vegetable origin, but there is very little data about their antimicrobial resistance phenotype and genotype or about the genetic lineages identified in this niche (Allydice-Francis and Brown, 2012; Kidd et al., 2012). For this reason, the purpose of this study was to analyze the occurrence of Pseudomonas spp. in food samples (meat and vegetables), and to characterize the antimicrobial resistance phenotype and genotype of the recovered isolates.
Materials and Methods
Bacterial isolation
From June 2011 to July 2012, 62 meat samples were analyzed for Pseudomonas spp. isolation (24 obtained in butcheries and 38 in supermarkets). The origins of these meat samples were as follows (number of samples): chicken (20), pig (19), beef (11), turkey (8), sheep (2), and mixed meat (2). After homogenizing 0.3 g of meat in 3-mL saline solution, 100 μL was inoculated in peptone water and incubated at 37°C during 48 h. Next, 100 μL of the suspension was streaked onto Cetrimide-agar plates that were incubated at 37°C during 24 h.
In July 2012, 59 vegetable samples were analyzed (35 tomatoes and 24 peppers): 13 of them obtained in local supermarkets, 23 in greengrocers, and 23 directly collected in orchards. All vegetables were loose and unprocessed. After homogenizing in a stomacher 25–30 g of vegetables and 100 mL of peptone water, 100 μL of this suspension was streaked onto Cetrimide-agar plates and incubated at 37°C during 24–48 h.
One or two colonies per meat or vegetable sample, presumptive of being Pseudomonas, were selected, and were identified by classical biochemical techniques (Gram stain, triple sugar iron, catalase, and oxidase reactions) and confirmed by a molecular method, using 16S rRNA gene fragment polymerase chain reaction (PCR) amplification (Lane, 1991), sequencing, and comparison with the GenBank database.
Antibiotic susceptibility
Susceptibility testing to 15 antipseudomonal agents (ticarcillin, piperacillin, piperacillin-tazobactam, ceftazidime, cefepime, aztreonam, imipenem [IPM], meropenem [MEM], doripenem [DOR], gentamicin, tobramycin, amikacin, ciprofloxacin, norfloxacin, colistin) was performed by the disc diffusion method following the guidelines of the Clinical and Laboratory Standards Institute (CLSI, 2013). Minimum inhibitory concentration (MIC) of IPM and MEM was performed by agar dilution method (CLSI, 2013) and MIC of DOR by Etest method (manufacturer's recommendations, bioMérieux). Extended-spectrum-β-lactamase (ESBL), metallo-β-lactamase (MBL), and class A carbapenemase phenotypes were determined by double-disc synergy tests (Jarlier et al., 1988; Lee et al., 2001; Doi et al., 2008). P. aeruginosa ATCC 27853 was used as a control strain.
Clonal relationship and molecular typing of isolates
The clonal diversity of our isolates was determined by pulsed-field gel electrophoresis (PFGE) of genomic DNA digested with SpeI enzyme. Bacterial DNA embedded in agarose plugs was prepared as previously described (Kaufmann, 1998). The PFGE conditions used were 2 ramps at 6 V cm–1, at 14°C and with pulse time ranging from 5 s to 15 s during 10 h and from 15 s to 45 s during another 10 h. DNA profiles were analyzed by the BioNumerics software 2.0 (Applied Maths, Belgium) choosing the Dice coefficient. Tolerance and optimization were set at 1.1% and 0.0%, respectively.
Multilocus sequence typing (MLST) was performed for P. aeruginosa strains by PCR and bidirectional sequencing of seven housekeeping genes as was described in the P. aeruginosa PubMLST website database (
Characterization of integrons and porin OprD
The presence of integrons was studied by PCR, and subsequent sequencing, to amplify the encoding genes for integrase of types 1, 2, and 3 (intI1, intI2, intI3), the 3′ conserved segment in class 1 integrons (qacEΔ1-sul1), and their variable regions (Sáenz et al., 2004).
Mutations in oprD gene were analyzed in P. aeruginosa strains by PCR, sequencing and comparison with P. aeruginosa PAO1 reference strain (GenBank accession no. AE004091) (Wolter et al., 2004; Gutiérrez et al., 2007).
Results
Isolates of Pseudomonas spp.
Two samples of meat (3%, turkey and beef) and 13 of vegetables (22%, 7 green peppers and 6 tomatoes) contained Pseudomonas spp. (Table 1). A total of 20 isolates were identified from the 15 positive samples, and were classified in 8 different species (number of isolates): P. aeruginosa (5), P. putida (5), P. nitroreducens (4), P. fulva (2), P. mosselii (1), P. mendocina (1), P. monteilii (1), and Pseudomonas sp. (1). The five isolates of P. aeruginosa were recovered from three samples of vegetable origin, and from one meat sample (Table 1).
Sequence type and oprD gene only were analyzed in the five P. aeruginosa strains.
Susceptible: This strain was susceptible to all 15 antipseudomonad antibiotics; TIC, ticarcillin; ATM, aztreonam; FEP, cefepime; IPM, imipenem; MEM, meropenem.
Reference porin OprD of P. aeruginosa PAO1 strain (GenBank accession no. AE004091).
Mutations detected in each OprD pattern:
Both strains were recovered from the same green pepper sample.
New allelic combinations: ST1455 (acsA15, aroE5, guaA11, mutL3, nuoD58, ppsA42, trpE9) and ST1456 (acsA11, aroE13, guaA109, mutL5, nuoD1, ppsA1, trpE47).
Both strains were recovered from the same green pepper sample.
Both strains were recovered from the same green pepper sample.
Both strains were recovered from the same tomato sample.
Both strains were recovered from the same green pepper sample.
Clonal relationship and molecular typing
All 20 isolates showed different PFGE patterns. P. aeruginosa strains were typed by MLST in five different sequence types: ST17, ST270, ST800, and two new allelic combinations that were registered in the MLST database as ST1455 and ST1456.
Resistance profiles
The Pseudomonas spp. strains recovered in this study showed the following resistance percentages: ticarcillin (85%), aztreonam (30%), cefepime (10%), IPM (10%), and MEM (5%), while they were susceptible to the other tested antimicrobial agents. Only one strain (Ps149), from meat origin, was susceptible to all tested antimicrobial agents. None of them harbored class A carbapenemase, MBL, or ESBL phenotypes. Table 2 shows the MIC values of carbapenem agents.
MIC, minimum inhibitory concentration; IPM, imipenem; MEM, meropenem; DOR, doripenem.
Presence of integrons and characterization of protein OprD
One IPM-resistant P. aeruginosa strain (Ps158) harbored a 3′ conserved segment-lacking class 1 integron, regulated by a weak promoter (PcW). All PCRs performed to detect the gene cassettes behind the intI1 gene were negative in this intI1-positive strain. The remaining 19 Pseudomonas isolates lacked integron structures.
Regarding protein OprD in the five P. aeruginosa strains, different amino acid changes were identified that were classified in three groups (Table 1). Four strains were identified in groups 1 and 2 containing several mutations in porin OprD. A new insertion sequence, designated ISPa47 by IS Finder (
Discussion
Humans, animals, foods, and the environment have been described as different reservoirs of bacteria harboring antibiotic resistance genes that could be transferred or mobilized to human pathogens (Rolain, 2013). Moreover, the extensive use and even the misuse of antimicrobial agents in clinic, animal production, and agriculture could be a way to select and disseminate these human pathogens (Rolain, 2013). Among them, P. aeruginosa is one of the most relevant opportunistic human pathogens, although there are also reports that show clinical cases caused by environmental Pseudomonas, such as P. mendocina (Nseir et al., 2011), P. fulva (Almuzara et al., 2010; Seok et al., 2010), P. mosselii (Giani et al., 2012) or P. monteilii (Bogaerts et al., 2011). In our study, 20 Pseudomonas isolates have been recovered from raw meat (3%) and vegetables (22%). The occurrence of P. aeruginosa contaminating tomatoes or green pepper products in our study (5%) was lower than those percentages obtained analyzing salads or other vegetables in previous works (64.5%, 44%, or 19%) (Wright et al., 1976; Correa et al., 1991; Allydice-Francis and Brown, 2012). In those studies, lettuce, chicory, carrot, and watercress yielded the highest frequencies of isolation. According to our results, and previously reported suggestions (Wright et al., 1976), the vegetables cultivated in more contact with the soil may be contaminated easily with Pseudomonas from soil, fertilizers, or water used for irrigation.
Antibiotic susceptibility testing revealed that all the Pseudomonas spp. isolated in this work were susceptible to aminoglycosides, fluoroquinolones, and colistin, and low resistance rates were observed for carbapenems and cefepime (≤10%). These results contrast with the frequent reports of multidrug-resistant P. aeruginosa isolated from clinical samples, and the increasing descriptions about carbapenem-resistant and MBL-producing isolates (Rojo-Bezares et al., 2014). In our study, only two IPM-resistant strains were found, and none of them was a MBL producer. Regarding the Ps159 strain, the resistance to IPM was associated with the loss of porin OprD by insertion of the new ISPa47 characterized in this study (GenBank KC502912). The inactivation of oprD gene expression by insertion sequence elements, leading to an increase in carbapenem resistance, has been previously reported only among clinical isolates (Wolter et al., 2004; Evans and Segal, 2007; Ruiz-Martínez et al., 2011a; Diene et al., 2013; Rojo-Bezares et al., 2014). To our knowledge, this is the first description of an environmental isolate resistant to IPM due to the oprD disruption by the presence of the ISPa47 element. On the other hand, the IPM-resistance in the Ps158 strain could not be determined in this work. The amino acid changes detected in its porin OprD were previously reported in carbapenem-susceptible P. aeruginosa isolates recovered from human clinical samples and from fecal samples of healthy humans (Ocampo-Sosa et al., 2012; Estepa et al., 2014). Patterns 1 and 2 of alterations in protein OprD might not cause porin loss; therefore, other resistance mechanisms (such as overexpression of efflux systems or AmpC) must be involved in the carbapenem resistance.
Several studies have associated antimicrobial resistance of P. aeruginosa with the presence of mobile genetic elements. The prevalence of class 1 integrons is high in clinical P. aeruginosa isolates (Ruiz-Martínez et al., 2011b; Martínez et al., 2012; Odumosu et al., 2013). In our study, class 1 integron was found in only 1 of the 20 Pseudomonas isolated (1 of the 5 isolated P. aeruginosa strains). The presence of integrons in these isolates is of great concern because these genetic elements are capable of spreading and capturing multidrug resistance gene cassettes. In addition, the dissemination of P. aeruginosa clones also favors the spread of antimicrobial resistance worldwide. In this sense, there are high-risk clones worldwide disseminated in several hospitals (such as ST235, ST111, and ST175) and normally associated with carbapenem-resistant strains and/or MBL producers (García-Castillo et al., 2011; Cabot et al., 2012). In our work, five different sequence types (ST17, ST270, ST800, ST1455, and ST1456) have been detected. ST1455 and ST1456 are new sequence types in which the two IPM-resistant strains were ascribed. According to the MLST database, the ST270 contains P. aeruginosa from water, soil, and clinical samples, and the ST800 from sputum samples. The ST17, detected in the P. aeruginosa Ps153 strain, is known as “clone C” and has been previously associated with cystic fibrosis and other non–cystic fibrosis patients (Kidd et al., 2012).
Conclusions
In conclusion, Pseudomonas spp. contaminating meat and vegetables samples were found in this study. The recovered Pseudomonas spp. belonged to many different clones and showed diverse antimicrobial susceptibility patterns. Fresh vegetable contamination can occur in the field by contaminated soil, by exposure to contaminated water (e.g., by crop irrigation and application of pesticides), or by deposition of human or animal feces. Also, raw meat could be contaminated due to animal, human, or environmental Pseudomonas at the slaughterhouse level, or during transformation or commercialization processes.
The antimicrobial resistance dissemination among bacterial populations is an increasing problem worldwide. Antibiotic resistance in Pseudomonas isolates, mainly P. aeruginosa and P. putida species, recovered from food is of particular concern because Pseudomonas are opportunistic human pathogens, and the food chain might be a vehicle of transmission of resistance genes to humans.
For all those reasons, not only consumers should be cautious with the manipulation and cooking of raw or fresh foods, but also surveillance throughout the food production and consumption chain is needed to detect emerging resistance phenotypes.
Footnotes
Acknowledgments
Vanesa Estepa had a predoctoral fellowship from the Universidad de La Rioja, Spain (grant number FPI-UR-09/16599009), during the experimental work for this study. This work was partially supported by the Instituto de Salud Carlos III of Spain (project FIS PI12/01276), and by the Ministerio de Economía y Competitividad of Spain (project SAF2012-35474) and Fondo Europeo de Desarrollo Regional (FEDER).
Disclosure Statement
No competing financial interests exist.
