Background
Salmonella enterica (
S. enterica) consists of more than 2600 different serovars, which, based on different antigens, are subgrouped into typhoidal and non-typhoidal
Salmonella (NTS) serovars. These serotypes vary greatly in their virulence, natural reservoirs and susceptibility to antimicrobials [
1]. Typhoidal
S. enterica serotypes
Typhi and
Paratyphi A and
B infect humans as their exclusive reservoirs, causing outbreaks of life-threatening typhoid fever [
2]. Outbreaks of typhoid fever are mainly restricted to low-income countries, while they represent less than 1 % infections in the developed countries. In contrast, non-typhoid salmonellosis usually gives rise to acute but self-resolving gastroenteritis, with no need of antimicrobial therapy; life-threatening invasive infections may occur in vulnerable patients [
3].
In 2015, the European food and safety authority (EFSA) reported that the investment of the European Union in
Salmonella control measures is yielding noticeable results, with only 20.4 confirmed cases per 100,000 population. Nevertheless, non-typhoidal salmonellosis remains the major cause of foodborne zoonotic outbreaks [
4].
The increasing number of antibiotic-resistant
Salmonella spp. against first line antimicrobial agents (aminopenicillins, trimethorpim-sulfametoxazole and chloramphenicol) intensified the empirical use of fluoroquinolones and third generation cephalosporins [
5‐
7]. Noticeably, these classes of antibiotics have been recommended for the treatment of
Salmonella infections in animals [
8]. In the last years, retail food, livestock and companion animals have been the main sources of NTS transmission [
9,
10].
Genetic determinants conferring resistance to beta-lactams and cephalosporins embedded in widely spreading plasmids have been extensively characterized in other
Enterobacteriaceae, such as
E. coli and
K. pneumoniae. The detailed genetic analysis of circulating plasmids helped containing MDR outbreaks [
11,
12]. Thus, monitoring serovar distribution and antimicrobial resistance of
S. enterica in humans may be important in controlling the infection.
The Microbiology and Virology Unit of the Padua Teaching Hospital, National Reference Centre for enteropathogenic bacteria and antimicrobial resistance epidemiology for the Northeast Italy, has begun in 2009 a strict surveillance of circulating non-pansusceptible Enterobacteriaceae, among which Salmonella spp.
Here we report the phenotypic and genotypic analysis of Salmonella isolates infecting human patients between January 2011 and December 2012 in Northeast Italy, analyzed at the Microbiology and Virology Unit of the Padua Teaching Hospital. We diagnosed and characterized multidrug resistance elements in two non-typhoidal S. enterica serovars, Rissen and monophasic
typhimurium, and in one typhoidal serovar, Paratyphi B. The resistance determinants were embedded in plasmids with the ability to transfer between bacterial species. Two S. enterica serotype Enteritidis samples were characterized by colistin resistance.
Methods
Characterization of Salmonella strains: serotyping and antimicrobial susceptibility testing
A total of 734 unique enteropathogenic isolates, obtained exclusively from human patients, were received and characterized by the Microbiology and Virology Unit of the Padua Teaching Hospital. Samples were collected between January 2011 and December 2012, according to the current version of the Declaration of Helsinki. Putative
Salmonella containing samples were plated on selective xylose lysine deoxycholate agar (XLD agar, Becton–Dickinson Italia, Milan, Italy) and colonies from pure cultures were characterized using a VITEK
® 2 automated system (bioMérieux, Marcy l’Étoile, France). The isolates identified as
Salmonella spp. were subjected to serotyping by slide agglutination, according to the Kaufmann-white classification [
13].
Antimicrobial susceptibility testing was performed on all confirmed
Salmonella isolates, by the disc diffusion (Kirby-Bauer) method according to EUCAST criteria [
14]. The following antibiotics (μg/disk) were tested on Mueller–Hinton agar (Difco Laboratories, Detroit, MI), according to Enter-Net specifications: nalidixic acid (30), ampicillin (10), cefotaxime (30), ciprofloxacin (5), chloramphenicol (30), gentamicin (10), kanamycin (30), streptomycin (10), sulphonamide (300), tetracycline (30), trimethoprim (5), ampicillin-clavulanic acid (20-10), cefalotin (30) and sulfamethoxazole-trimethoprim (23.75–1.25) (BD BB™ Sensi-Disk™, Becton–Dickinson Italia, Milan, Italy). Moreover, the disk diffusion test was performed to detect Extended-spectrum beta-lactamases (ESBL) (BD BB™ Sensi-Disk™, Becton–Dickinson Italia, Milan, Italy).
Colistin susceptibility was determined using Sensititre ARIS® 2X (Trek Diagnostic Systems, Thermo Scientific Inc., Milan, Italy). Escherichia coli ATCC25922 and E. coli ATCC35218 were used as control strains.
Identification of ESBL and AmpC determinants: plasmid transfer by transformation and conjugation experiments
Plasmid DNA was obtained from pure bacterial cultures performing phenol/chloroform extraction. Resistance genes (bla
TEM, bla
SHV, bla
CTX-M, and AmpC) were detected by PCR as previously described [
15,
16]. Positive PCR products were sequenced in an ABI3130xl sequencer (Applied Biosystems, Thermo Scientific Inc., Milan, Italy) and sequences were aligned to those present in GenBank using BLAST (
http://blast.ncbi.nlm.nih.gov/Blast.cgi).
Plasmids containing PCR-confirmed resistance determinants were introduced into chemo-competent E. coli cells (SCSI, Agilent Technologies Italia, Milan, Italy). Transformants were plated on selective Mueller–Hinton (cefotaxime 2 mg/L or ampicillin 100 mg/L) (Becton–Dickinson & Co., DifcoTM) agar plates. The presence of plasmids was confirmed by PCR targeting the whole sequence of the bla
TEM/CTX/CMY gene; PCR products were treated with exosap and fully sequenced (ABI3130xl sequencer, Applied Biosystems). Microbial identification and antimicrobial susceptibility testing on transformants were performed at a VITEK® two automated system (bioMérieux, Marcy l’Étoile, France).
For conjugation experiments, one colony of each donor (S. enterica serovars: Rissen, monophasic S. typhimurium 1,4 [5],12:i- and Paratyphi B) and the recipient strain E. coli J53 (F− met pro Azir Colr) were cultured separately under weak agitation in LB broth at 37 °C. Conjugation was performed incubating different dilutions (from 1:100 to 1:10) of donor suspensions with overnight culture of the recipient strain under low agitation. Mating assays were carried out at room temperature from 1 h to overnight depending on the donor strain. The transconjugants were selected on agar plates containing colistin (8 mg/L) and cefotaxime (2 mg/L) or ampicillin (100 mg/L). Microbial identification and antimicrobial susceptibility testing on transformants and transconjugants were performed by VITEK® two automated system (bioMérieux, Marcy l’Étoile, France).
Plasmid PCR-based replicon typing (PBRT)
All plasmids were analyzed by PCR-based replicon typing (PBRT) targeting the replicons of the major plasmid families occurring in
Enterobacteriaceae [
17]. Each vector was assigned to a specific Inc group. Plasmid typing was also performed in the obtained transformants and transconjugants.
Amplification and analysis of pmrA and pmrB genes
Bacterial total DNA was obtained by heat shock from pure culture suspension.
PmrA and
pmrB genes, the major regulators of LPS modifications in
S. enterica, were amplified as previously described [
18]. PCR amplicons were sequenced by end-primers in an ABI3130xl sequencer (Applied Biosystems, Thermo Scientific Inc., Milan, Italy) and the obtained sequences were compared to the corresponding sequences in the NCBI database (
http://www.ncbi.nlm.nih.gov/nucleotide).
Discussion
Non-typhoidal salmonellosis has not been hitherto considered a major public health risk in developed countries, such as the European Union member states. Up to date, only few epidemiology studies on
Salmonella isolates infecting humans has been published so far [
19,
26,
27].
In the present study, we presented the Italian epidemiology of Salmonella isolates infecting humans between 2011 and 2012, both in terms of serovar distribution and in terms of antimicrobial resistance patterns.
Our investigation highlighted that pathogenic
Salmonella counted for the majority of all enteropathogenic pathogens (47.8 %) diagnosed by our laboratory, National Reference Centre for Enteropathogenic Bacteria for the Northeast Italy. In the 2-year period of the study, more than 11 different serotypes were identified:
monophasic S.
typhimurium 1,4 [5],12:i:- and
S. typhimurium were the prominent circulating serovars, retrieved in almost 30 and 27 % of all samples, respectively. In contrast, the EFSA report on circulating zoonotic agents in 2013 indicated that
S. enteritidis was the most prevalent
Salmonella serovar detected in human samples [
28].
Our epidemiological data are supported by recent works demonstrating that
S.
typhimurium and
monophasic S. typhimurium 1,4 [5],12:i- have extremely high survival rates in the environment [
29], can infect a broad range of species [
30] and have enhanced their ability to cause salmonellosis in humans [
19]. Moreover, these were the most common serotypes diagnosed in serious foodborne salmonellosis outbreaks in both humans and animals in 2012 [
31]. Thus,
S.
typhimurium and
monophasic S. typhimurium 1,4 [5],12:i- have to be considered emerging predominant serotypes in humans all around Europe.
In addition, our study pointed out that 61.6 % of analyzed
Salmonella isolates were resistant to at least one antibiotic, with 13.3 % of samples classified as multidrug resistant. In accordance with our data, EFSA reported increasing multidrug resistance in
Salmonella infecting humans, where
monophasic S. typhimurium 1,4 [5],12:i- serotype exhibited the highest rate of multidrug resistance at the European level [
32].
Despite the relatively low resistance rate against the clinically important third generation cephalosporins, molecular analysis of strains displaying ESBL/AmpC-producing phenotypes proved that the isolates bore worrisome plasmid-embedded resistance determinants, able to move through bacterial species: AmpC, such as CMY-2, and ESBL, such as TEM-52 and CTX-M-15. The three determinants have been more commonly associated with other members of the
Enterobacteriaceae family [
33]. Notably, the identified genes have been recently reported in human outbreaks of salmonellosis all around the world [
19,
20,
34], associated with other
Salmonella serovars and embedded in plasmids distinct from the ones traced in this study.
So far, ESBL have been rarely described in
S. paratyphi B at the European level and have been never reported at the Italian level [
35,
36]. Moreover,
Salmonella serotype
Rissen is a widespread contaminant of pigs and pork products [
37] and up to date the presence of CMY-like cephalosporinases in this serovar had been reported only in South Korea in 2002 [
38,
39]. Our data highlight the possible future spreading of resistance to first line antimicrobials through this serovar.
An additional worrisome feature described here is the detection of two isolates of
S. enteritidis displaying resistance to colistin. In Europe,
S. enteritidis is reported to be one of the most frequent serotypes diagnosed both in animals and in humans. No other studies detected and characterized, up to date, colistin-resistant isolates belonging to this serovar. Up to 2013, the highest proportions of resistance among
S. enteritidis isolates were observed for nalidixic acid (19.5 %) and ampicillin (11.0 %), while resistance to third generation cephalosporins was generally detected at low levels [
32]. The spreading of colistin-resistant
S. enteritidis could worsen the clinical treatment of life-threatening cases of salmonellosis.
Authors’ contributions
IF carried out the genotypic assays, performed the analysis and interpretation of data and drafted the manuscript, SB performed the phenotypic analysis, EDC performed the phenotypic analysis and data collection, SNR conceived of the study, participated in its design and coordination and revised the manuscript; GP participated in the study coordination. All authors read and approved the final manuscript.