A total of 307 Salmonella isolates were obtained from a laboratory-based surveillance network for enteric pathogens coordinated by the Bacteriology Division from the Central Public Laboratory of the Rio Grande do Sul State (Laboratório Central do Estado do Rio Grande do Sul, LACEN-RS). All these samples were originated from clinical samples collected from cases of salmonellosis in hospitalized patients or ongoing emerging cases in the southernmost Brazilian State, Rio Grande do Sul. Epidemiological data (collection date/season of the year, local/city of collection, clinical sample) were also gathered from the 13 regional clinical laboratories (Coordenadorias Regionais de Saúde – CRSs) of the complete network in Rio Grande do Sul state (with 18 laboratories) and provided by LACEN-RS.

Clinical samples were first plated on Salmonella/Shigella solid media (Probac, São Paulo, Brazil). Presumptive Salmonella colonies were then cultured in triple sugar iron (TSI) (PlastLabor, Rio de Janeiro, Brazil), urea, citrate and lysine iron agar (LIA) slants (Probac, São Paulo, Brazil) with aerobic incubation at 37 °C for 24 h. Salmonella suspect colonies were confirmed by agglutination test with polyvalent sera O and H commercial antigens (Probac, São Paulo, Brazil). Complete antigenic characterization and serotype identification were carried out in the National Reference Laboratory for Cholera and Enteric Diseases, Oswaldo Cruz Institute (FIOCRUZ, Rio de Janeiro, Brazil). The somatic (O) antigens were determined by slide agglutination tests and flagellar (H) antigens were determined using a microplate agglutination technique produced by the laboratory. Serotypes were assigned according to the Kauffmann–White–Le Minor (KWL) scheme.18,19

In vitro susceptibility testing to twelve antibiotics was performed by agar disk diffusion according to the recommendations of the Clinical and Laboratory Standards Institute (CLSI).20 The twelve antibiotics used were from nine different antimicrobial categories: aminoglycosides (streptomycin – Str, gentamicin – Gen), cephalosporins (ceftazidime – Caz, cefoxitin – Cfo), penicillins (ampicillin – Amp), quinolones (nalidixic acid – Nal, ciprofloxacin – Cip), tetracyclines (tetracycline – Tet), phenicols (chloramphenicol – Clo), folate pathway inhibitors (trimethoprim/sulfamethoxazole – Tm/Sut), nitrofurans (nitrofurantoin – Nit) and carbapenems (imipenem – Imp). After 24 h of incubation at 36 °C ± 1 °C, clear zones (circles) were identified in the non-resistant samples. The diameter of the inhibition zone was interpreted as resistant, intermediate and susceptible according to CLSI guidelines.20Escherichia coli ATCC 25922 was used as the control bacteria strain. Isolates were classified as MDR if they were resistant to three or more antimicrobial categories according previously described.21

DNA extraction of the bacterial isolates was performed by boiling method as described previously.22 A real-time PCR targeting invA gene was carried out for Salmonella generic detection and a duplex real-time PCR targeting fliA-IS200 and safA genes was used for the specific detection of the serotypes Typhimurium and Enteritidis, respectively. Primers/probes sequences and amplification conditions were previously described.23,24 All experiments were carried out in the StepOnePlus™ Real-Time PCR Systems (Thermo Fisher Scientific, Waltham, Massachusetts, USA) with cycling conditions of 95 °C for 3 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 60 s. Negative and positive controls were added in all PCRs.

Data were analyzed using the Statistical Package for Social Sciences (SPSS, version 23.0, Chicago, IL). The Pearson's chi-squared test of independence and/or Fisher's exact test were used to verify possible statistical differences between the qualitative variables. Residual adjustment analysis was used to identify the resistance profile of the serotypes to the antimicrobials evaluated. All statistical tests were two-sided and p values < 0.05 were considered as statistically significant.

All the 307 isolates were submitted to the invA real-time PCR and presented positive result to Salmonella. The number of isolates by year was not constant and ranged from 34 in 2010 to 72 in 2012. In addition, there was a similar rate of the Salmonella isolates in Spring (1st September to November 30th) (89, 29%), Summer (1st December to 28th February) (87, 28.3%) and Autumn (1st March to 31th May) (80, 26.1%), but it was observed a reduction of the number of isolates in Winter (1st June to 31th August) (51, 16.6%). All Salmonella isolates were originated from 44 different sub-regions from the Rio Grande do Sul state, covering all geographic macro regions and the most important urban centers in southern Brazil (Fig. 1). The highest number of isolates was obtained in the metropolitan area from the state's capital Porto Alegre (CRS1) with a total of 117 (57.7%) samples, while there was only one isolate in two regions (CR13 and CRS17). No Salmonella isolates were obtained in five macro-regions (CRS7, CRS11, CRS12, CRS15 and CRS16).

Fig. 1.

Salmonella serotypes most commonly reported (number of isolates) in clinical samples of the different macro-regions (CRSs) from southern Brazil. Each circle represents the frequency of the serotypes Enteritidis, Typhimurium and others in the different macro-regions (CRSs).

(0.45MB).

The total 307 isolates were from 289 different patients and overall socio-demographic information is demonstrated in Table 1. Salmonellosis frequency was higher in adults with 45 years old or more (105, 36.3%) than in other ages, but there was also an important number of children below five years (54, 18.7%), children and adolescents (34, 11.8%) and adults with less than 45 years old (89, 30.8%). Salmonella was isolated from several different human clinical specimens and classified into non-enteric (blood, urine, abscesses, biological fluids, and aspirates) and enteric (stool) samples according to the source of each sample. Among non-enteric samples 79 (64.2%) were considered invasive because they were isolated from the blood and other clinical fluids such as cerebrospinal, pleural, synovial and ascites (Table 1).

A total of 287 isolates from S. enterica subsp. enterica and two isolates from S. enterica subsp. houtenae were detected in the six-year period spanning this study. TS serotypes Typhi and Paratyphi were not detected among the isolates from the subspecies enterica, so all 289 isolates were NTS. The scenario is clearly dominated by serotypes Typhimurium and Enteritidis (193, 66.7%; p < 0.01). Typhimurium was the most frequent one with 140 (48.4%; p < 0.01) isolates, while Enteritidis was detected in 53 (18.3%) of the samples. All these isolates also presented positive result only for the respective serotype in the duplex real-time PCR targeting fliA-IS200 and safA genes.

Among the remaining isolates of the subspecies enterica, 83 (28.7%) were from 21 different serotypes and 11 (3.8%) were not completely serotyped (Table 2). All the 83 isolates of the 21 serotypes and seven samples with incomplete antigenic formula presented negative results in the duplex real-time PCR, but the remaining four isolates presented PCR positive result for serotype Typhimurium (all of them with the antigenic formulae O:4,5). With these four isolates, the prevalence of Typhimurium rises to almost half (144, 49.8%) of all samples. These serotypes were isolated in all regions of the state (Fig. 1).

The overall number and frequency of Enteritidis isolates was higher than Typhimurium (and also than the others serotypes) in 2010. The reverse situation (first Typhimurium, second Enteritidis) occurred in 2011. Typhimurium was significantly more frequent than Enteritidis in the other following 4 years (p < 0.01) (Fig. 2A).

Fig. 2.

Epidemiology of human salmonellosis in southern Brazil, 2010 to 2015. (A) Salmonella isolates from all samples. (B) Salmonella isolates from hospitalized patients. The isolates were classified into three serotypes groups: Typhimurium, Enteritidis and others.

(0.25MB).

Salmonella isolates of this study were from sporadic cases in hospitalized patients or associated with ongoing salmonellosis emerging cases. Most isolates (244; 84.4%) were from hospitalized patients and occurred in all geographic macro regions of southern Brazil. Typhimurium was more frequent than all other serotypes, with 135 (55.6%; p < 0.01) isolates, followed by Enteritidis (33; 13.6%). The sum of the other serotypes reached 75 (30.9%) isolates. Interestingly, the overall number and frequency of Typhimurium isolates was much higher than any other serotypes in all years, ranging from 40% in 2010 to 67.7% in 2015 (Fig. 2B).

The remaining 45 isolates were related to 23 regional food-borne salmonellosis emerging cases (9 in 2010, 6 in 2011, 3 in 2012, 4 in 2014 and 1 in 2015). These emerging cases were more frequent in Summer (12, 52.2%), but they were also reported in the three other seasons: Spring (6, 26.1%), Autumn (4, 17.4%) and Winter (1, 4.3%). There were 21 emerging cases with the identification of only one serotype: Enteritidis (14; 60.4%), Typhimurium (3, 14.3%), Infantis (1; 4.8%), Braenderup (1; 4.8%), Corvallis (1; 4.8%) and Johannesburg (1; 4.8%). Noteworthy, there was one important emerging outbreak caused by serotype Infantis with eleven (24.4%) isolates, while all the other ones had only two to four isolates. In one case, two Salmonella serotypes were detected (Agona and Typhimurium).

A total of 280 (91.4%) NTS isolates were submitted to antimicrobial susceptibility with 12 antibiotics and classified into susceptible and resistant. Sixty-eight (24.3%) were sensitive to all of them, while the other 212 (75.7%) isolates presented resistance to one or more antibiotics. Among these last isolates, 108 (38.6%) were resistant to one or two antimicrobials categories, while the other 104 (37.1%) were resistant to three or more drugs and were considered MDR (Table 2). In an individual evaluation of each antibiotic, it is noteworthy the high frequency of the resistance to Nal (167, 59.2%), Tet (106, 37.6%), Amp (99, 35.1%), Nit (98, 34.2%), Str (92, 32.6%), Gen (54, 19.1%), Cip (44, 15.6%), Clo (43, 15.2%) and Tm/Sut (33, 11.8%). Two other antibiotics showed resistance rates below 5%: Cfo (8, 2.8%) and Caz (4, 1.4%).

In the analysis of the antibiotic resistance according to the serotype, most Typhimurium isolates presented resistance to three or more antibiotics, while the majority of the Enteritidis isolates presented resistance to two antibiotics and most of the other serotypes were susceptible to all tested antibiotics (Table 2). In the analysis of the resistance to different antibiotics categories, Typhimurium was the serotype with the highest number of MDR isolates (83, 72.2%; p < 0.01). Interestingly, twelve (14.4%) Typhimurium isolates presented the multi-resistant profile ACSSut (resistance to ampicillin, chloramphenicol, streptomycin, sulfamethoxazole, and tetracycline). In addition, ten (8.7%) Enteritidis isolates and 22 (19.1%) of the other serotypes also were MDR (Table 2). An overall view of the antibiotic resistance according to serotypes (Typhimurium, Enteritidis and others) is demonstrated in Fig. 3.

Fig. 3.

Antimicrobial resistance rate of different antimicrobial agents in nontyphoidal Salmonella in southern Brazil, 2010 to 2015. The isolates were classified into four serotypes groups: Typhimurium (blue), Enteritidis (red), others (green), and with incomplete antigenic formulae (purple). Antimicrobial resistance was classified according the following antibiotics classes: aminoglycosides (Str, streptomycin, Gen, gentamicin), cephalosporins (Caz, ceftazidime, Cfo, cefoxitin), penicillins (Amp, ampicillin), quinolones (Nal, nalidixic acid, Cip, ciprofloxacin), tetracyclines (Tet, tetracycline), phenicols (Clo, chloramphenicol), folate pathway inhibitors (Tm/Sut, trimethoprim/sulfamethoxazole), nitrofurans (Nit, nitrofurantoin) and carbapenems (Imp, imipenem).

(0.15MB).