1 Introduction

Salmonella enterica enterica belongs to the most common enteropathogenic bacteria worldwide, causing diseases varying from mild gastroenteritis to severe systemic infections. The type of disease is determined to a great extent by the virulence characteristics of the involved Salmonella strain, as well as by the individual host defense mechanisms. A variety of virulence factors have been described for Salmonella, some of them showing a broad distribution, while others are present in a limited number of serovars or even strains [1, 2].

Among them, the type III secretion system (TTSS) encoded in the Salmonella pathogenicity island (SPIs) is considered a “quantum leap” in the evolution of Salmonella as a pathogen [1, 3]. The TTSS is required for induction of proinflammatory responses, invasion of intestinal epithelial cells, induction of cell death in macrophages, and elicitation of diarrhea [4–7].

This type III system allows the bacteria to secrete and inject bacterial toxins (effector proteins) directly into the cytosol of host cells. The translocated proteins are the actual mediators of the virulence phenotypes associated with the SPIs type III secretion system.

Salmonella effector proteins SopE, SopE2, SopB, SopD, SptP, and AvrA are encoded in SPIs of the TTSS apparatus [4, 7–13].

In the last 2 decades, the emergence and spread of antimicrobial- resistant pathogens, including Salmonella, have become a serious health hazard worldwide. The routine practice of giving antimicrobial agents to domestic livestock for preventing and treating diseases, as well as for growth promotion, has contributed to the emergence of antibiotic-resistant bacteria that are subsequently transferred to humans by the food chain [14–16]. Of particular concern is the increasing association of human infections with multidrug-resistant Salmonella strains [17, 18]. There are undoubtedly many factors associated with the emergence of resistance. An understanding of these factors is crucial if we want to limit the spread of resistance. A significant percentage of clinical drug resistance and consequent therapeutic failure in Salmonella is known to be associated with the formation of biofilms [19–23].

In the present study, the distribution of the effector protein genes sopB, sopD, sopE, sopE2, avrA, and sptP located on the SPIs was analyzed as such, and evaluated as well, with respect to biofilm formation and the patterns of antimicrobial resistance in various serovars of Salmonella enterica enterica strains isolated from children with diarrhea. The effector proteins that we studied were SopB (activating Akt/protein kinase B, a signal transduction kinase that can exert prosurvival effectors) [7, 9, 13, 24], SopE (playing a role in the efficient entry of the bacterium into the cells) [4, 9, 13], SptP (inhibiting the expression of NF-κB-dependent genes and potentially affecting the balance of pro- and antiapoptotic factors in the cell) [25], AvrA (inhibiting the activation of NF-κB) [9, 26], SopD (contributing to inflammation and fluid secretion, enhancing membrane fission, and promoting macropinocytosis) [27, 28], and SopE2 (activating Rho-GTPase leading to cytoskeletal rearrangements and bacterial entry) [13, 29].

Although several studies about Salmonella enterica enterica deal with the presence of virulence genes, the antimicrobial susceptibility, and the ability of these bacteria to form biofilms, there are no reports about any possible association of these properties with each other [10, 11, 30–33]. There is only one previous study investigating the association of virulence genes with antimicrobial resistance in Salmonella enterica enterica serovars other than S. enteritidis and S. typhimurium [12]. However, the results are not comparable, as this report—except for SopE and SopB—referred to different genes, and it did not include the resistance of biofilm forms.

2 Methods

Salmonella enterica enterica strains (n = 194) were collected from children with gastroenteritis, either hospitalized or attending the outpatient clinic of Penteli Children’s Hospital Athens, Greece, aged from 1–14 years. According to the history, there was no treatment with antibiotics before admission to the hospital in any of the patients. The isolation and serological identification of S. enterica enterica were performed by conventional methods. An agar plate with a pure culture was prepared for each strain, and a single colony from each plate was resuspended in 1 ml of sterile water in a microfuge tube and stored at −80 °C until use for the DNA extraction.

Biofilm formation determination and antimicrobial resistance of the planktonic and biofilm forms of the strains were performed as reported previously [19]. Biofilm formation was detected using silicone disks (Folio C6 0.25 mm, NOVATECH; New Biotechnology for Life, Z.I. Athélia III, Voie Antiope 13705, La Ciotat Cedex, France). The silicone disks were cut into similar sizes (4–5 mm), weighed (25–30 mg) in a spacer construction (invented in-house), placed in tubes, weighed on a scale, and left overnight under UV irradiation for sterilization. Tubes containing 2.5 ml trypticase soya broth (TSB) were inoculated with Salmonella strains and incubated for 72 h at 30 °C. The contents were then poured off, and the tubes were washed three times with distilled water and air-dried in a laminar flow for 24 h. The tubes containing the silicone disks with the attached bacteria were weighed once more, and the difference in weight showed the presence of biofilms.

The antimicrobial susceptibility of the planktonic bacterial forms was performed by determination of the minimum inhibitory concentration (MIC). The MIC was determined using two methods: (1) the automatic VITEK 2 system (bioMérieux SA, 69280 Marcy-l’Etoile, France) and (2) the standard broth dilution method according to guidelines of the Clinical Laboratory Standards Institute (CLSI, Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria That Grow Aerobically, 8th edition: Approved Standard M7-A8; 29:2. CLSI, Wayne, PA, USA). For the antimicrobial susceptibility determination, the antibiotics tested were ampicillin, cotrimoxazole, ciprofloxacin, moxifloxacin, gentamicin, cefuroxime, cefotaxime, co-amoxiclav, and imipenem.

For the determination of the antimicrobial susceptibility of the biofilm forms, silicone disks coated with the biofilm-forming Salmonella strains were prepared in tubes as described above, omitting the last step (air-drying). Serial dilutions of the antimicrobials in Mueller-Hinton broth, corresponding to the concentrations used for the MIC determination of the planktonic forms, were prepared and poured into the silicone disk-containing tubes. The antimicrobial-containing tubes were then incubated at 35 °C for 24 h. The growth of planktonic bacteria was visualized by the development of turbidity in the medium. The Minimal Inhibitory Concentration for Bacterial Regrowth (MICBR) from the biofilm was defined as the lowest concentration showing no growth in the medium as observed by a complete clarity.

As described previously, the biofilm forms showed significantly increased antimicrobial resistance versus the planktonic bacteria. The respective results of planktonic versus biofilm forms were: ampicillin 12.8 versus 84.4 %, gentamicin 0 versus 89.9 %, co-amoxiclav 0 versus 51.4 %, cotrimoxazole 0.9 versus 63.3 %, cefuroxime 7.4 versus 63.3 %, cefotaxime 0.9 versus 23.8 %, imipenem 0 versus 7.3 %, ciprofloxacin 0 versus 2.8 %, and moxifloxacin 0 versus 2.8 % [19].

All the strains were screened for the presence of sopB, sopE, sopD, avrA, spt-P, and sopE2 using a PCR assay, as described previously [4, 34]. The oligonucleotide primer pairs, corresponding to the six virulence gene fragments, were run in separate reactions (Table 1).

Table 1 PCR primers and conditions for detecting sopB, sopE, sopD, avrA, spt-P, and sopE2
Full size table

DNA extraction from Salmonella culture was performed with the boiling method and by means of the InstaGene Matrix commercial reagent, according to the manufacturer’s instructions (Bio Rad Laboratories, 2000 Alfred Nobel Dr., Hercules, CA, USA). The PCR mixture contained 25 μl JumpStart REDTaq ReadyMix, 15 pmol/μl of Forward primer, 15 pmol/μl Reverse primer, 17 μl H2O, and 5 μl DNA Salmonella. Amplification was carried out in an iCycler (Bio Rad Laboratories, Hercules, CA, USA). Subsequent electrophoretic analysis of 15 μl of PCR amplified samples in 2 % agarose gel in TBE (Tris-borate-EDTA) buffer was carried out at a constant voltage of 85 V for 1 h 50 min. Gels were stained with ethidium bromide and were observed under UV light. A digital image of the gel was captured in a computer, and the amplification patterns were evaluated by visual examination of inverted gel pictures.

The statistical package SPSS (version 17) was used. Fisher’s exact test was used, which is a statistical significance test used in the analysis of contingency tables where sample sizes are small.

3 Results

Of the 194 Salmonella enterica enterica clinical isolates from infected children with diarrhea, 19 different serotypes were identified (Table 3). S. enteritidis (144/194) was the predominant serotype followed by S. typhimurium (27/194). From the total of the isolated strains, 56 % (109/194) were positive for biofilm formation. The biofilm-forming strains showed increased antimicrobial resistance compared to the planktonic bacteria as reported previously [19]. Electrophoretical analysis showed that the effector protein gene sptP was detected in 99 % of the strains, followed by sopB in 98.5 %, sopD in 97.4 %, avrA in 97.4 %, sopE2 in 77.8 %, and sopE in 76.3 %, respectively. Ten different genotypes were present in our study (Table 2). At least one virulence gene associated with SPI-1 was present in all Salmonella isolates. All six genes were detected in 144 (74.3 %) strains, 5 were present in 9 (4.6 %), 4 in 34 (17.5 %), 3 in 4 (2 %), 2 in 1 (0.5 %), and 1 in 2 (1 %) (a detailed presentation is given in Table 3).

Table 2 Genotypes detected among the Salmonella enterica enterica serovars with respect to the six effector protein virulence genes investigated
Full size table
Table 3 Distribution of the effector protein virulence genes among the various Salmonella enterica enterica serovars and the respective genotypes
Full size table

Statistical analysis showed that there was a significant association regarding the presence of the virulence genes: sopB associated with sopD (P < 0.002) and sopE2 (P < 0.010), sopD associated with sopB (P < 0.002), sptP (P < 0.001), and avrA (P < 0.005), and sopE associated with avrA (P < 0.001) and sopE2 (P < 0.000). The presence of sptP was associated with sopD (P < 0.001), of avrA on sopD (P < 0.005), sopE (P < 0.001) and sopE2 (P < 0.000). Finally, sopE2 was associated with the presence of sopB (P < 0.010), sopE (P < 0.000), and avrA (P < 0.000).

The ability to form biofilm was statistically independent of any of the six genes. Although several genes have been reported to be associated with biofilm formation in Enterobacteriaceae so far, the genetic basis of biofilm production seems not to be restricted to a specific gene, but it might be a result of multigene function and expression [35, 36].

Because of the small number of strains belonging to each of the ten different genotypes found (Table 4), three main groups were formed based on the dispersion rate of the six virulence genes. This classification facilitated the assessment of any possible association of the genes with the other phenotypic features (biofilm formation, antimicrobial susceptibility of the planktonic and the biofilm forms).

Table 4 Antimicrobial resistance rates of the biofilm forms of Salmonella enterica enterica serovars among the various genotypes
Full size table

These groups were as follows: group I: strains of genotype A (presence of all the six genes), group II: strains of genotype E (absence of only sopE and sopE2), and group III: all the other gene combinations.

None of the three groups was statistically associated with the ability of form biofilm. Strains lacking sopE and sopE2 were more resistant to all the antimicrobials. There was a significant positive association of genotype E (group II) with resistance of the planktonic forms to ampicillin (13 strains out of 34, P < 0.001), a phenomenon not found in the other two groups, including either strains bearing all six genes or strains missing one or more genes like in group III. Another statistically significant association was found between the biofilm forms of group II (18 out of 21, P < 0.001) and resistance to cotrimoxazole. It seems that the exclusive absence of sopE and sopE2 might be related to enhanced antimicrobial resistance in general and specifically a significant resistance to ampicillin in the planktonic forms and to cotrimoxazole in the biofilm forms.

4 Discussion

In our study, the absence of SopE and SopE2 correlated with enhanced antimicrobial resistance. As these two genes functionally facilitate the bacterial entry to the host cell, which is a very important step for bacterial host cell invasion, through the induction of inflammation [7, 9, 13, 37], the lack of the genes might lead to the emergence of antimicrobial-resistant strains through natural selection in order to counterbalance this “handicap.”

There are few reports dealing with this association on a molecular basis, but the emergence of more virulent and antibiotic-resistant Salmonella strains is a topical public health problem. Some investigations are directed towards the possible formation of responsible plasmid cointegrates [2, 12], which enable the transfer of the genes among various strains or serovars.

The presence of the genes in the bacterial chromosome of the strains included in our study was not associated with biofilm formation. However, in a previous report it was proved that the expression of these genes was significantly associated with the formation of biofilms and cell clumps [10]. This is a point needing further investigation as many of our strains bore the genes without forming biofilms (Table 4).

The reported incidence of the effector protein virulence genes in Salmonella varies depending on the serovar and the origin [11, 12, 31–33, 38]. The rates reported in our study and elsewhere are in agreement with each other when considering clinical isolates [11, 12, 32]. The majority of our serovars encoded all six effector protein virulence genes (Tables 2, 3) [12]. So far, it seems that, in clinical cases, the contribution of these genes to pathogenicity is questionable, as many other virulence factors are described to be associated with Salmonella enterica enterica pathogenicity [39–41]. In our study, some strains bearing a single gene (either sptP or avrA) were still capable of causing gastroenteritis. SptP and avrA are both associated with the inhibition of the expression and activation of the transcription factor NF-κB (nuclear factor kappa beta) [9, 25, 26]. This factor plays an important role in the immune system of the host, regulating the expression of cytokines, the inducible nitric oxide synthase, several growth factors, inhibitors of the apoptosis and effector enzymes, as well as in the development and the activity of a number of tissues including the central nervous system [42, 43]. perhaps further investigations should be focused on the importance of sptP and avrA in the emergence of Salmonella enterica infection in humans.