Abstract
A total of 50 Salmonella enterica strains were isolated from clinical samples from 2009 to 2012 and analyzed for the presence of virulence genes found in SPI-1, SPI-2, and plasmids. The distribution and frequency of the antimicrobial resistance genes and plasmids were revealed, and pulsed-field gel electrophoresis (PFGE) patterns were investigated. Five genes were identified from the seven strains with resistance or intermediate resistance to ampicillin: blaSHV-1 (present in six strains), qnrS1 (present in five strains), blaTEM-1 (present in three strains), blaCTX-M-1 (present in one strain), and qnrB1 (present in one strain). One trimethoprim–sulfamethoxazole-resistant strain was positive for sulI but negative for sulII. In addition, we detected TEM-1 and qnrS1 in one strain; SHV-1 and qnrS1 in two strains; TEM-1, SHV-1, CTX-M-1, and qnrS1 in one strain; TEM-1, SHV-1, and qnrB1 in one strain; and SHV-1 and sulI genes in one strain together. Plasmid-based replicon typing assay revealed that all 50 strains carried FIIS, 13 carried I1, 1 carried I2, 4 carried P, 1 carried A/C, and 4 carried X1 replicon. PFGE was used to type 46 of the 50 strains and classify them into 22 major groups, 33 pulsotypes, and 8 major clusters. All strains carried all the virulence genes of interest on both Salmonella Pathogenicity Islands 1 and 2 and plasmids suggested high potential for pathogenicity. All antimicrobial-resistant strains contained at least one of the resistance genes of interest, confirming a phenotype–genotype association in antimicrobial resistance.
Introduction
S
Virulence determinants of Salmonella strains are located in Salmonella Pathogenicity Islands 1 and 2 (SPI-1 and SPI-2, respectively) and virulence plasmids (pSTV). SPI-1 is associated with the invasion of the gastrointestinal epithelium, while SPI-2 is linked with the survival and replication in phagocytic cells, which are required for the later stages of the infection. 5 Plasmidial virulence genes facilitate the systemic spread and colonization of deeper tissues. 1
The use of antimicrobials in livestock, companion animals, and humans has led to the emergence of multidrug-resistant strains. Horizontal gene transfer enables the transmission of resistance genes among bacteria strains through mobile elements, particularly plasmids. Thus, monitoring antimicrobial resistance genes and the distribution of plasmids among Salmonella strains is important to reveal antimicrobial resistance patterns, which will shape efficient antimicrobial therapy.6,7
Molecular typing methods have been widely used to establish relatedness between strains, evaluate outbreaks, and determine their source. Pulsed-field gel electrophoresis (PFGE) has been considered as the gold standard technique for Salmonella molecular typing.1,2,8 Thus, in this study, we investigated the genetic factors that affect the virulence of Salmonella. The goal was to identify the antimicrobial resistance genes and plasmids that confer resistance to several antimicrobials in multidrug-resistant strains, as well as to determine the clonal relationship between strains.
Materials and Methods
Bacterial strains
A total of 50 Salmonella enterica strains belonging serogroups D (n = 36), C (n = 8), and B (n = 6) were isolated from clinical samples from patients admitted to the Trakya University Health and Research Application Centre Central Laboratory Microbiology Department between 2009 and 2012. Bacterial identification tests were conducted using conventional methods and the automated system VITEK2 (BioMerieux, France) using gram negative test cards. In addition, Antimicrobial susceptibility tests were performed with VITEK2 AST cards and strains were grouped as “susceptible,” “intermediate,” or “resistant” according to VITEK2 Systems 7.01 software and CLSI Document M100-S20. Strains were grouped using multiplex polymerase chain reaction (PCR) and slide agglutination with Salmonella polyvalent and group-specific Salmonella Poly A-I+Vi, Salmonella O:2, O:4, O: 6,7,8, O:7,8, O:8, O:9, O:3,10,15, and O:1,3,19 antisera (Plasmatec, United Kingdom). Multiplex PCR was performed using six sets of primers targeting the O-antigen synthesizing gene regions in serogroups A, B, C1, D, and E, which are all commonly found in clinical isolates. Salmonella internal control primers targeting the oriC region were included in each test. Salmonella standard strains (S. Paratyphi A, S. Typhimurium, S. Choleraesuis, S. Enteritidis, S. Newport) obtained from the Turkish Public Health Institution were used as positive controls in O-grouping. 9
DNA extraction
A single colony from each strain was suspended in 50 μL of ultrapure water. The suspension was heated at 95°C for 10 min and centrifuged at 14,000 rpm for 10 min. Thirty microliters of the supernatant was used as a DNA template for PCR.
Virulence genes
Each strain was analyzed by PCR to identify the invA, sipA, sipD, sopB, sopD, and sopE2 virulence genes located in SPI-1, the ssaR and sifA genes located in SPI-2, and the spvB and prot6E genes in pSTV. Table 1 lists the sequences, predicted amplicon sizes, and references of the primers of interest.
Primers Sequences and Amplicons of the Salmonella Virulence Genes Studied
Antimicrobial resistance genes
According to antimicrobial susceptibility results, seven strains showed resistance or intermediate resistance to ampicillin. These were analyzed for the presence of TEM-1, SHV-1, and CTX-M-1 beta lactam resistance genes and qnrA1, qnrB1, and qnrS1 plasmid-mediated quinolone resistance genes. One strain that proved resistant to trimethoprim–sulfamethoxazole was also assayed for the presence of sulI and sulII sulfonamide resistance genes. Table 2 presents the sequences, predicted amplicon sizes, and references of these primers.
Primers Related to Antimicrobial Resistance Genes Used in This Study
M = A or C; H = A or C or T; Y = C or T.
All amplifications were conducted using 10 pmol of each primer (Biomatik), 0.5 U of Taq polymerase (Thermo Scientific), and 3 μL of DNA template in a reaction mixture with a total volume of 25 μL. Amplicons were separated in a 2% agarose gel (Prona, Spain) by unidirectional electrophoresis in 1 × TBE buffer (Gibco, Thermo Fisher Scientific) and visualized by staining with ethidium bromide. Fragment size was determined by comparison with 100 bp or 1 kb DNA ladders (Thermo Fisher Scientific) according to the predicted size of the amplicons.
Plasmid-based replicon typing
Plasmid-based replicon typing (PBRT kit-Diatheva, Italy) was used to investigate the distribution and frequency of the plasmids responsible for antimicrobial resistance in the Salmonella strains. The kit was used to investigate the presence of 25 replicons that may confer resistance to several antimicrobials in Enterobacteriaceae strains by providing eight multiplex PCR mixtures (M1–M8). DNA was extracted as described above. Amplifications were conducted in 25-μL volumes containing 21.8 μL of the reaction mixture, 3 μL of DNA, and 0.2 μL of Taq polymerase. PCR was performed as previously described.6,10 Fragments were separated by unidirectional electrophoresis in 2.5% agarose gel, as described above. The replicons from the different multiplex PCR mixtures are listed in Table 3.
Distribution of the Replicons Investigated by Plasmid-Based Replicon Typing According to Polymerase Chain Reaction Mixtures and Amplicon Sizes
PCR, polymerase chain reaction.
Pulsed-field gel electrophoresis
Agarose blocks of 50 strains were prepared using a previously described protocol with the following modifications. 2 Half of each plug was digested in 5 μL of XbaI (10 U/μL), 10 μL of 10 × XbaI buffer, and 85 μL of ultrapure water overnight at 37°C. The restriction fragments were separated by electrophoresis in 0.5 × TBE buffer at 14°C for 18 hr using the CHEF DR III drive module (Bio-Rad Laboratories, Hercules, CA) with pulse times of 6 to 22 sec. The gels were stained with EtBr (1 μg/mL), and DNA band patterns were analyzed with GelCompar II software (Applied Maths, Sint-Martens-Latem, Belgium) to determine the relatedness of the strains. A similarity dendrogram was constructed by the unweighted pair group method with arithmetic mean (UPGMA) method using the Dice similarity coefficient.
Results
Serogrouping demonstrated that 50 strains belonged to serogroups D (36), C (8), and B (6). According to antimicrobial susceptibility tests that were conducted with an automated system, seven strains were found to be resistant to at least one antimicrobial agent (Table 4). The PCR results showed that all 50 strains carried all investigated virulence genes located on SPI-1, SPI-2, and pSTV. All antimicrobial-resistant strains carried at least one of the resistance genes of interest.
List of Strains Resistant to At Least One Antibiotic and Minimum Inhibitory Concentration Values
The MIC values of strains that are accepted as resistant (R) according to CLSI 2011 criteria are written in bold.
Intermediate resistant (I).
MIC, minimum inhibitory concentration.
Among the seven strains that have resistance or intermediate resistance to ampicillin, blaSHV-1 was found in six (85.7%), blaTEM-1 was found in three (42.8%), and blaCTX-M-1 was found in one (14.3%) strain. Moreover, blaTEM-1 and blaSHV-1 co-occurred in one strain and blaTEM-1, blaSHV-1, and blaCTX-M-1 in another. The same seven strains were investigated for the presence of qnrA1, qnrB1, and qnrS1 genes; qnrS1 was found in five strains (71%), while qnrB1 (14%) was found in one strain. The strain resistant to trimethoprim–sulfamethoxazole was positive for sulI and negative for sulII.
PBRT assay demonstrated that all 50 strains carried FIIS (Salmonella virulence plasmid), 13 carried I1, 4 carried P, 4 carried X1, 1 carried A/C, and 1 carried I2. The PBRT results revealed that five out of the seven resistant strains with at least one antimicrobial resistance gene have at least one plasmid in addition to the Salmonella virulence plasmid (FIIS). In comparison with the susceptible strains, the resistant strains had a significantly higher rate of having another plasmid in addition to FIIS (p < 0.05 to p < 0.01).
PFGE was used to type 46 out of the 50 strains. These 46 strains were classified into 22 major groups, 33 pulsotypes, 8 major clusters (belonging to the 9.1, 9.3, 10.1, 11.1, 11.2, 11.4, and 15.1 pulsotypes), and 25 specific clusters with similarity above 85%. The genetic relatedness of the strains ranged from 35% to 100%. One Escherichia coli strain was added to the study, and its relatedness to the Salmonella strains was below 30%.
The XbaI PFGE types of the isolates are shown in Figure 1. Strains in the same serogroup showed closer relatedness to each other compared with strains from other serogroups. Strains in serogroup D displayed 35% relatedness to the strains in serogroups B and C, and strains from serogroups B and C were 45% similar. All strains and their antimicrobial resistance genes, plasmids, and pulse types are listed in Supplementary Table S1 (Supplementary Data are available online at www.liebertpub.com/mdr).

Dendrogram representing genetic relatedness among Salmonella strains based on PFGE fingerprints. PFGE, pulsed-field gel electrophoresis.
Discussion
Salmonellosis is a major public health concern responsible for high morbidity and large economic and social burdens. Salmonella is transmitted primarily by contaminated food and water and is estimated to cause over 93.8 million infections and 155,000 deaths per year worldwide. Use of antimicrobials in livestock, companion animals, and humans has led to the emergence of antimicrobial resistance in bacterial pathogens and has in some cases limited therapeutic options. These problems reveal the importance of epidemiological studies that investigate the virulence properties and antimicrobial susceptibility of strains in the community, as well as to help identify points of contamination, develop control programs, and monitor the results.1,11–13
The pathogenicity of the Salmonella strains is associated with the virulence genes located on its chromosomal pathogenicity islands and plasmids. SPI-1 is responsible for the invasion of the intestinal epithelium and diarrheal symptoms during localized gastroenteritis, whereas SPI-2 is required for the survival and replication in phagocytes, which are crucial for the later stages of the infection and for systemic spread. 5 Although most of the virulence factors are located on the SPIs, a few responsible for the colonization of extraintestinal organs (e.g., liver, spleen) and disseminated infections are carried on Salmonella pSTVs.
In our study, PCR results revealed that all strains carried all virulence genes located on SPI-1, SPI-2, and plasmids. This suggests the high pathogenic potential of the clinical Salmonella isolates in Turkey. Similar results were obtained in previous studies conducted on strains isolated from humans, pigs, and poultry.1,2,14
Antimicrobial therapy is not usually recommended to treat gastroenteritis with Salmonella, but it is necessary in cases of severe infections (e.g., bacteremia, meningitis) or for immunosuppressed patients. However, emerging antimicrobial resistance is an important challenge to treatment, which can originate from the use of antimicrobials in livestock, companion animals, and humans.
Despite being susceptible to a majority of antimicrobials up to the 1960s, clinicians are encountering difficulties in treatment of gastroenteritis caused by Salmonella with third-generation cephalosporins due to extended spectrum beta-lactamase (ESBL) production of the strains. Resistance can be transferred between strains with plasmids and other mobile elements, and therefore, the genetic cause of resistance is important.15–18 Antimicrobial susceptibility patterns differ within the serotypes of Salmonella. Despite being the most common serotype in the community, S. enterica ser. Enteritidis is more susceptible than other serotypes isolated from clinical strains. While S. enterica ser. Typhimurium tends to develop resistance, S. enterica ser. Enteritidis strains are still susceptible to many antimicrobials.
Thus, to reveal serogroup–antibiotic susceptibility patterns, our study serogrouped all of the strains with polyvalent and group-specific antisera and compared the results to the VITEK2 antibiogram findings. In addition, two (33%) strains of serogroup B, two (25%) strains of serogroup C, and three (8%) strains of serogroup D were resistant to at least one of the tested antimicrobials, which demonstrated different levels of antimicrobial resistance development between the serogroups. More of the isolates collected in 2011–2012 (5/26) were resistant to at least one antimicrobial than those collected in 2009–2010 (2/24). In 2009–2010, all strains were susceptible to ampicillin, but in 2011–2012, five strains were found to be resistant, and one strain showed resistance to the combination of beta-lactam/beta-lactamase inhibitor and third-generation cephalosporins.
Strains that are resistant or have intermediate resistance to ampicillin were analyzed for the presence of TEM-1, SHV-1, and CTX-M-1 beta-lactam resistance genes. Among the seven strains, blaSHV-1 was found in six (85.7%), blaTEM-1 was found in three (42.8%), and blaCTX-M-1 was found in one (14.3%). Moreover, one strain carried all three genes. According to the PCR results, all resistant strains carried at least one beta-lactamase synthesizing gene, which supports the correlation between phenotype and genotype. Previous studies demonstrated close relationships between ESBL and plasmid-mediated quinolone resistance genes, highlighting the importance of revealing the prevalence of qnr-positive strains in the community. As ESBL and qnr genes usually cotransferred among the strains, qnr positivity may suggest also ESBL production. 19 Therefore, the presence of qnrA1, qnrB1, and qnrS1 genes was investigated in all seven strains. While qnrA1 was not detected in any of the strains, qnrS1 and qnrB1 were found in five strains (71%) and one strain (14%), respectively. Moreover, a trimethoprim–sulfamethoxazole-resistant strain that also carried SHV-1 was positive for sulI and negative for sulII. Two antimicrobial-susceptible strains were added to each PCR as controls and were negative for all resistance genes investigated.
Taken together, all antimicrobial-resistant strains carried at least one of the resistance genes of interest, indicating a phenotype–genotype association in antimicrobial resistance. Similar results on the phenotype–genotype relationship were obtained in previous studies on strains isolated from other geographic regions.7,15,20 Plasmid-mediated quinolone resistance genes were not found in any of the isolates collected in 2009–2010, but at least one qnr gene was found in those collected in 2011–2012. Also, SHV-1 was the only beta-lactamase gene found in the strains isolated in 2009–2010, but TEM-1 and CTX-M-1 were detected in addition to SHV-1 in strains from 2011 to 2012.
Horizontal gene transfer is the main mechanism for the spread of antimicrobial resistance determinants between different bacterial strains and often involves plasmids. Plasmid-coded ESBLs, carbapenemases, and methylases were identified in Enterobacteriaceae strains in previous studies. However, the plasmids are not found at the same frequency in all enteric bacteria, with some plasmids having high frequency in only some strains. Therefore, we used PBRT to reveal the distribution of the plasmids isolated from clinical Salmonella strains to establish their clonal relationships.6,10,21 We used eight multiplex PCRs to investigate the presence of 25 replicons that are commonly found in Enterobacteriaceae strains (HI1, HI2, I1, I2, X1, X2, L/M, N, FIA, FIB, FIC, FII, FIIS, FIIK, W, Y, P, A/C, T, K, U, R, B/O, HIB-M, and FIB-M). Six of them were detected in our clinical Salmonella strains (FIIS, I1, I2, P, A/C, and X1).
FII replicons belonging to the IncFII type plasmids were defined as virulence plasmids (FIIK, FIIY, and FIIS) in Klebsiella, Yersinia, and Salmonella species. SpvA, spvB, and spvC genes carried on IncFIIS plasmids contribute to the systemic phase of Salmonella infection.22–25 In our study, FIIS was found in all 50 strains, which indicates a high virulence potential of the clinical isolates in Turkey. Moreover, the spvB gene, which is carried on the IncFIIS plasmid, was detected in all isolates by PCR. That the same results were obtained with the PCR and PBRT assays proves the robustness and reliability of these methods
It has been demonstrated that some Salmonella strains carry other plasmids in addition to the virulence plasmid. Although most are associated with antimicrobial resistance, some afford different advantages to the bacteria, including lactose usage, phage conversion, and resistance to heavy metals.26,27 IncI1 plasmids are responsible for the synthesis of type IV pili in Shiga toxin-producing E. coli strains associated with ESBLs, particularly blaCTX-M-1 production.28–32 In this study, IncI1 was found in 13 strains, including 1 positive for blaCTX-M-1, 1 positive for blaSHV-1, and 1 positive for both blaTEM-1 and blaSHV-1, as well as 10 susceptible strains.
IncI2 that is also associated with beta lactamases codes two types of pili similar to IncI1. The blaCTX-M-55 and blaCMY-2 genes were found on IncI2 plasmids, indicating the role of IncI2 in the distribution of beta lactamases.33–36 In this study, IncI2 was found in one strain carrying blaTEM-1, blaSHV-1, blaCTX-M-1, and qnrS1.
IncP plasmids replicate in various hosts and encode several genes that confer resistance against different groups of antimicrobials, in addition to heavy metals and ammonium compounds that are found in disinfectants. In addition to pathogenic and opportunistic bacteria, IncP-derived plasmids have been isolated from soil, fertilizer, and waste water samples, suggesting that plasmids could provide selective advantages for bacteria and will likely contribute problems in infection control.35,36 Aminoglycoside, tetracycline, beta lactam, and sulfonamide resistance genes were previously identified on IncP plasmids.37–40 In this study, IncP was found in one isolate positive for blaSHV-1 and qnrS1, one positive for blaSHV-1, qnrS1, and sulI, and two phenotypically susceptible isolates. Since additional plasmids were found in resistant isolates, further sequencing studies are needed to unequivocally relate resistance genes and IncP.
IncA/C was initially isolated in the 1970s and is another plasmid group that acts as a reservoir for different antimicrobial resistance genes, including blaCMY-2 and sulI.41–43 Moreover, a blaNDM-1-positive New Delhi metallo-β-lactamase coding IncA/C variant was recently found in India. Its broad host range makes this plasmid group an even greater concern for public health.44,45 In this study, IncA/C was identified in one strain positive for blaSHV-1, sulI, and qnrS1.
IncX1 plasmids identified before the antimicrobial era are associated with resistance to olaquindox, quinoxaline, quinolones, and fluoroquinolones through the production of an efflux pump. In this study, IncX1 was found in four strains, including one possessing blaTEM-1 and qnrS1. Significantly, more of the resistant strains than the susceptible strains carried another plasmid in addition to the FIIS Salmonella virulence plasmid (p < 0.05 to p < 0.01). This suggests that plasmids play an active role in acquiring antimicrobial resistance and that these mobile elements likely help in the dissemination and transmission of resistance within strains.
While genetic relatedness ranged from 35% to 100% among the Salmonella strains after XbaI digestion, the relatedness to E. coli was below 30%. As expected, strains in the same serogroup showed closer relatedness to each other in comparison to the strains belonging to other serogroups. Strains in serogroup D showed 35% relatedness to the strains from serogroups B and C, while strains from serogroups B and C showed 45% similarity between them. Interestingly, the relatedness of two ampicillin- and levofloxacin-resistant strains from serogroup C was above 70%, but their closeness to the susceptible strain from serogroup C was ∼45%, which is the same as their relatedness to strains of serogroup B. Another striking result was the relatedness between the two susceptible strains of serogroup B, which was above 70%, while their closeness to the blaTEM-1, blaSHV-1, and qnrB1-positive strain of serogroup B was around 52%. The low relatedness among resistant strains and susceptible strains of the same serogroup suggests that the acquisition of antimicrobial resistance genes is effective in clonal associations.
Conclusion
All of the 50 strains investigated in this study carried a virulence genes of interest located on SPI-1, SPI-2, and plasmids, which is indicative of a high potential for pathogenicity. In addition, all antimicrobial-resistant strains carried at least one of the associated resistance genes, thereby supporting the phenotype–genotype association in antimicrobial resistance. Also, in comparison to the susceptible strains, significantly more of the resistant strains carried another plasmid in addition to the Salmonella virulence plasmid, which highlights the role of plasmids in acquiring antimicrobial resistance. The plasmids associated with antimicrobial resistance in the Salmonella strains identified in this study may complicate future treatment options for cases of salmonellosis due to horizontal gene transfer.
Footnotes
Disclaimer
All subjects gave their informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of Trakya University.
Acknowledgment
This study was supported by a grant (Project No. 2014-76) from Trakya University Research Fund (TUBAP).
Disclosure Statement
No competing financial interests exist.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
