Nontyphoid Salmonella gastroenteritis in pediatric patients from urban areas in the city of Mérida, Venezuela

Background: Nontyphoid Salmonella (NTS) infections are a frequent cause of self-limited diarrhoeal illness in healthy children that do not usually require antibiotic treatment. This study was conducted by analyzing the phenotypic and genotypic traits of NTS strains from pediatric patients with acute gastroenteritis living in urban areas in the city of Mérida, Venezuela. Methodology: Thirty-seven Salmonella strains (18 S. Enteritidis; 14 S. Typhimurium; 2 S. Java; 2 S. Saintpaul; 1 S. Infantis) were isolated from 243 stool specimens. These strains were biochemically identified and serotyped. Antimicrobial susceptibility was determined by the disc-diffusion assay. Genetic characterization included plasmid profiling, PCR detection of the spv region and inv genes, and IS200 typing. Results: Thirty (81.0%) of the Salmonella isolates were resistant to the antimicrobial tested. Of these strains, 17 (56.7%) were resistant to at least one antibiotic. Five resistance patterns were observed, of which the most frequently found was the single type (tetracycline, streptomycin or ampicillin). All the S. Typhimurium harbored plasmids, but only three large plasmids (60, 72 and 84 kb) yielded amplicons with a spvR specific primers. All the Salmonella serotypes showed the presence of an inv region. Eight distinct IS200 profiles could be detected among the 37 Salmonella strains studied. Conclusions: Predominant Enteritidis and Typhimurium serotypes, as well as serotypes Java, Saintpaul and Infantis, are circulating in the city of Mérida, Venezuela. Most of these strains are susceptible to first-line antibiotics but active monitoring of isolates for antimicrobial resistance is necessary. IS200 typing, applied in association with conventional methods, allowed the characterization of all isolates and suggested the presence of different infection sources.


Introduction
Nontyphoid Salmonella (NTS) is a common cause of gastroenteritis, which is usually a self-limited illness in healthy children that does not require antimicrobial treatment [1]. However, in invasive infections such as meningitis or septicemia, effective antibiotic therapy can be lifesaving [1,2].
As happens with most enteropathogens, there has been an alarming increase of resistance to different antimicrobial agents, and especially to those most commonly used in developing countries, such as ampicillin, trimethoprim/ sulfamethoxazole, chloramphenicol or tetracycline [2,3]. The spread of resistant Salmonella strains is a relevant issue for pediatricians, because few therapeutic options are available to treat infants and children [2,4]. More rapid identification of the pathogen and its antibiotic sensitivity pattern could potentially guide antibiotic choice, reducing unnecessary use of expensive second-line antibiotics for susceptible isolates and improving the outcome for multiresistant isolates [3,5,6]. Unfortunately, in Venezuela, limited financial and laboratory resources generally preclude the use of routine stool cultures and antimicrobial drug susceptibility testing to guide therapy. Hence, an empiric therapy is often employed without supporting data and is usually based on the clinical experience of the physician.
Moreover, epidemiological studies of NTS infections are hampered by the lack of adequate procedures. Standard methods, including phenotyping procedures, may not be discriminatory [7]. In recent years, molecular-based techniques, such as plasmid profile analysis, ribotyping, random amplified polymorphic DNA analysis, pulsed-field gel electrophoresis (PFGE), Rep-PCR and IS200 profiling, have shown to be useful methods for discriminating J Infect Developing Countries 2009; 3(1): 28-34. among isolates of Salmonella species [8][9][10]. IS200 is currently considered to be one of the most reliable typing procedures and a suitable epidemiological tool, because it is able to discriminate among strains from different geographic origins and also to relate strains isolated from a given area [8].
The aim of the present study was to determine serotype distribution and resistance patterns among NTS strains isolated from pediatric patients with acute gastroenteritis from urban areas in the city of Mérida, Venezuela, during November 2004 to November 2005. The study further aimed to evaluate the genotypic characteristics of NTS strains causing this infection.

Patients and bacterial strains
From November 2004 to November 2005, thirtyseven Salmonella strains (18 S. Enteritidis; 14 S. Typhimurium; 2 S. Java; 2 S. Saintpaul; 1 S. Infantis) were isolated from stool specimens obtained from pediatric patients with acute gastroenteritis from urban areas in the city of Mérida and admitted to the Instituto Autónomo Hospital Universitario de Los Andes (Mérida, Venezuela). The average age of the children studied was three years; 66% were female. The typical duration of illness was six days and cultures were usually obtained within 48 hours of onset. Freshly excreted whole stool specimens were collected from each patient, inoculated into Cary Blair medium (Oxoid Ltd., Basingstoke, UK), and transported to the laboratory within three hours after inoculation. Slides of fecal specimens were stained with Gram and methylene blue and were examined for fecal leukocytes by using light microscopy; specimens with > 5 leukocytes per high-power field were considered positive for inflammatory diarrhoea.
Isolation and biochemical identification of Salmonella species were performed using standard procedures. Salmonella strains were grouped with polyvalent antisera and serotyped based on somatic O and phases I and II flagellar antigens, by agglutination test with antisera (Bio-Rad and WHO Collaborative Centre for Reference and Research on Salmonella) as specified by the White-Kauffman-Le Minor scheme [11].

Antimicrobial susceptibility testing
All isolates were screened for resistance to 16 antibiotics by the disc diffusion method on Mueller-Hinton agar (Oxoid), according to the guidelines of the Clinical and Laboratory Standards Institute (CLSI) [12].

Plasmid analysis
Plasmid DNA was extracted by the alkaline denaturation method of Birnboim and Doly [13]. Plasmids were electrophoretically separated in 0.8% horizontal agarose gels at 100 V for three hours in Trisborate-EDTA buffer. Gels were stained with ethidium bromide and photographed under UV illumination. Reference plasmids (RP4, 54 kb; pIPI73, 126 kb) and the supercoiled DNA ladder (Invitrogen Life Technologies, Karlsruhe, Germany) served as size standards.

Detection by PCR of the spv region and inv genes
PCR was performed with the following sets of primer pairs: PG48 (CCCCGGGAATTCGCTGCATAAGGTCAGAAGG) and PG49 (CCCCGGGATCCAT-GGATTTCTTGATTAATAAA) for the spvR gene [14], and INVA-1 (ACAGTGCTCGTTTACGACCTGAAT) and INVA-2 (AGACGACTGGTACTGATCGAT-AAT) for the invasion gene invAE [15]. The reaction mixtures and cycling conditions of amplification were conducted as previously described [14,15]. The amplified products were separated by electrophoreses on 1.2% agarose gels and stained with ethidium bromide (0.5 mg/l); amplicons were visualized and photographed under UV illumination.

Extraction and digestion of cellular DNA
Genomic DNA was extracted as previously described [16] from exponential cultures. DNA was digested with the PstI (Promega, Madison, WI) restriction enzyme known to have no recognition site within the IS200 sequence.

IS200 profiling
Southern blotting of PstI-cleaved genomic DNA was performed as previously described [17]. The probe was generated by PCR amplification of a 557-bp internal fragment of the IS200 element, using the IS200-F CCTAACAGGCGCATACGATC and IS200-R ACATCTTGCGGTCTGGCAAC primers [18]. Preparation of the dUTP-11 fluorescein-labeled PCRgenerated IS200 probe, prehybridization, and hybridization were carried out using the ECL random prime labeling and detection systems (version II; Amersham Biosciences) in accordance with the manufacturers. All hybridizations were performed under highly stringent conditions. Membranes were exposed to X-ray film (ECL Hyperfilm; Amersham Biosciences).

Calculation of discriminatory power
The discriminatory power of the IS200 profiling was determined by calculating the indices of discrimination (D) according to Hunter and Gaston [19].

Statistical analysis
Variables in NTS-susceptible and NTS-resistant isolates were compared using chi-squared test. A value of p 0.05 was considered significant.

Results
A total of 243 pediatric patients with acute gastroenteritis from urban areas in the city of Mérida, Venezuela, were included in the study. Cultures of fecal specimens from 37 children (15.2%) yielded Salmonella organisms. S. Enteritidis and S. Typhimurium were the most frequently isolated serotypes, accounting for 18 (48.7%) and 14 (37.8%) isolates, respectively. The remainder was a distribution of the other serotypes, in which S. Java and S. Saintpaul accounted for two (5.4%) cases each, and S. Infantis was isolated in one case (2.7%). The remaining patients were excluded from microbiologic analysis because a different enteropathogen was identified in the fecal specimens (Shigella spp. 25.1%, Aeromonas spp. 21.1%, Campylobacter spp. 14.5%, enterotoxigenic Escherichia coli 9.2% and others 4.6%) or because specimens yielded more than one pathogen (10.3%). The relationship between clinical findings and susceptible or resistant Salmonella strains isolated from pediatric patients with acute gastroenteritis are shown in Table 1. Of the 37 strains of Salmonella isolates, 30 (81.0%) were resistant to at least one of the antimicrobial agents tested, while seven (18.9%) were susceptible to all antibiotics tested. Diarrhoea and presence of blood and leukocytes in feces were the clinical and laboratory characteristics most frequently observed in the patients. Also, the nutritional conditions of most patients were within normal values and approximately one third of them consumed some type of antibiotic at least two weeks before this study. There was no significant difference in clinical characteristics between the NTS-susceptible and NTS-resistant isolates. The findings obtained in the antimicrobial susceptibility assays (Table 2) showed the following results: two (11.1%) of the S. Enteritidis strains and all S. Java, S. Saintpaul and S. Infantis strains were susceptible to the antibiotics tested, while all of the S. Typhimurium strains were resistant to at least one antibiotic. Five resistance patterns were observed, but the most frequent was the single-type resistance (Tet, Str or Amp). Tetracycline was the resistance marker most frequently observed, both in strains resistant to one or two antibiotics (Tet+Str or Tet+Amp). Regardless of serotypes of Salmonella isolated, these strains were susceptible to amoxicillin/clavulanic acid, cephalothin, cefuroxime, ceftazidime, nalidixic acid, ciprofloxacin, chloramphenicol and trimethoprim/ sulfamethoxazole (data not shown).
Of the 37 Salmonella strains studied, 30 carried 9.5 to 84 kb plasmids, and of these strains, eight carried other smaller plasmids (2 to 6.5 kb) ( Table 3). None of the susceptible Salmonella strains were found to have plasmids. All the S. Typhimurium harbored plasmids, but only three large plasmids (60, 72 and 84 kb) yielded amplicons with the spvR-specific primers and were then assigned to the Salmonella virulence plasmid group. All the serotypes of Salmonella isolated showed a 457 bp band of amplification confirming the presence of the inv region (Table 3). Eight distinct profiles could be detected within the 37 Salmonella strains studied (Table 3 and Fig. 1).   The number of copies of the IS200 element was 11 in profile D and 2 in G. Profiles A, B and C belonged to S. Enteritidis strains. Of these profiles, B was the most frequent, occurring in eight of the 18 isolates of S. Enteritidis, whereas profiles A and C were observed in each of five strains of this serotype. The IS200 element was found in restriction fragments of approximately 10 kb and 6.4 kb and were a common feature in these three profiles (A, B and C). Profiles D and E belonged to S. Typhimurium strains. Nine of the 14 serotype Typhimurium strains were included in profile D. S. Java, S. Saintpaul and S. Infantis displayed 3 profiles (F, G and H), and these strains had 4, 2 and 3 copies of IS200, respectively.
The discrimination index, determined by combining serotyping and IS200 typing of the 37 NTS isolated, was 0.856.

Discussion
In this study, the most frequent serotypes of Salmonella causing acute gastroenteritis in children were S. Enteritidis (48.7%) and S. Typhimurium (37.8%). These results are in line with those described by Hohmann [1] and de Jong and Ekdahl [20], in which the dominant serotypes causing food-borne gastroenteritis in humans are S. Enteritidis and S. Typhimurium.
Bloody dysentery, the most pathognomonic feature of salmonellosis, was observed in the majority of children studied. None of the patients developed extraintestinal complications.
The emergence of antimicrobial resistance within Salmonella species has been reported worldwide [2]. In the present study, a strong prevalence of antibioticresistant Salmonella was observed (81.0%). Similar to previous reports [21][22][23], S. typhimurium was the most frequent resistant serotype. Although S. Enteritidis is usually reported sensitive to antibiotics [2], we found that this serotype was resistant in 17 (88.9%) cases. Regardless of Salmonella serotype, resistance to tetracycline was the most commonly observed. This antibiotic is not routinely used in pediatric patients in our country. Tetracycline is one of the drugs largely used in animals, reinforcing the commonly accepted view that the use or abuse of veterinary antibiotics is the leading cause of drug resistance spreading among S. enterica strains [4,22]. It is noteworthy that the Enteritidis and Typhimurium serotypes remained susceptible to cephalosporins, chloramphenicol, and other antibiotics that are usually considered as first-line treatment for severe or invasive infections caused by Salmonella. In contrast with other studies [6,[24][25][26][27], S. Java, S. Saintpaul and S. Infantis were fully susceptible to the antibiotics tested. Although different susceptibility patterns may arise from a number of factors, it is possible that differences in the susceptibility patterns among the studied serotypes of Salmonella can be ascribed both to different reservoirs and to the sources in which these strains were isolated (sporadic cases of gastroenteritis). Hence, a continuous monitoring of both antimicrobial resistance and deliberate use of antimicrobial agents in animals and humans are essential.
The virulence of Salmonella is linked to a combination of chromosomal and plasmid factors. The chromosomally located invasion gene invA, involved in cellular invasion, was present in all studied serotypes of Salmonella. These results agree with those of other authors who have found that this gene is a good target for detecting Salmonellae [15,28,29]. Plasmid profiling is a traditional method used for epidemiological studies in Salmonella strains within specific serotypes [9]. However, in this study, all resistant serotypes carried plasmids of different sizes and the frequency of an additional small plasmid was low. None of the susceptible strains harbored plasmids. Nevertheless, further studies will be necessary to know if the resistance observed is due to these plasmids. Because the results from plasmid profiling did not represent a discriminatory method for subtyping these strains, the plasmids detected were of little, if any, epidemiological importance, as reported in previous studies [30,31].
The true prevalence of virulence plasmids among natural isolates of Salmonella is unknown [31]. In this study, only three resistant Typhimurium serotypes carried virulence plasmids (spvR positive). Some studies support that plasmid-encoded spv genes enhance the ability of certain serotypes of Salmonella to produce severe extraintestinal disease [33,34], but these were not observed in the group of patients studied. It is possible that one or more factors, related to the genetic background of these isolates, is required to obtain virulence in strains carrying the spv-encoding virulence plasmid.
Molecular typing has enhanced the understanding of the epidemiology of Salmonella infections by increasing our knowledge of the genetic relationships among strains [7][8][9][10]. The molecular approaches used in this work allowed a more sensitive characterization of these isolates. Determination of the number of copies and localization of IS200 fragment insertion showed a variety of profiles in the studied strains, including those belonging to the same serotype (Enteritidis and Typhimurium). Enteritidis serotype strains showed two common bands in IS200 profiles, indicating that these particular strains might be clonally related. However, resistance patterns and IS200 profiles did not show a clear correspondence. The IS200 typing seems to provide a more reliable molecular marker for epidemiological studies, because it clearly discriminated among strains of different serotypes isolated from pediatric patients with gastroenteritis.
In addition, the presence of various serotypes with different phenotypic and genotypic characteristics suggests that the infection originated from multiple sources. It is possible that the presence of Enteritidis and Typhimurium serotypes is due to human-to-human transmission, while Java, Saintpaul and Infantis serotypes could be introduced into the community by food.
In summary, this study showed that predominant Enteritidis and Typhimurium serotypes, as well as Java, Saintpaul and Infantis are circulating in urban areas in the city of Mérida, Venezuela, and also that most of these strains are susceptible to first-line antimicrobial agents used to treat severe salmonellosis in children. Nevertheless, active monitoring of serotypes isolated for antimicrobial resistance is necessary because of the public health implications of a potential spread of resistance clones. IS200 typing, applied in association with conventional methods, allowed a better characterization of all isolates and suggested the presence of different infection sources. Furthermore, our findings indicate that enteric infections by nontyphoidal salmonellae continue to pose a major threat to children in Mérida, Venezuela. Identification of sources of human infection, early clinical and microbiological diagnosis, appropriate treatment, education about basic food hygiene, and implementation of a strict sanitary control of the food industry are all necessary to reduce the prevalence of salmonellosis.