Salmonella enteritidis

The following bacterial species were used: Bacillus subtilis (ATCC 10707), Enterobacter cloacae (human isolate), Escherichia coli (ATCC 0157:H7), Micrococcus flavus (ATCC 9341),Proteus mirabilis (human isolate), Pseudomonas aeruginosa (ATCC 27853), Salmonella enteritidis (ATCC 13076), S. epidermidis (ATCC 12228) S. typhimurium (ATCC 13311) Staphylococcus aureus (ATCC 25923). The antibacterial assays were carried out by the disc-diffusion [25 (link)] and microdilution method [26 ,27 (link),28 (link)] in order to determine the antibacterial activity of oils and their components against the human pathogenic bacteria. The bacterial suspensions were adjusted with sterile saline to a concentration of 1.0 × 10⁵ CFU/mL. The inocula were prepared daily and stored at +4 °C until use. Dilutions of the inocula were cultured on solid medium to verify the absence of contamination and to check the validity of the inoculum.

Salmonella Typhimurium strain 81.23500, Salmonella Enteritidis strain CVD SE and Salmonella Dublin strain 06-0707 were used to develop the multiplex PCR. Twenty-four control strains which came from the Salmonella Reference Laboratory of the Centers for Disease Control and Prevention (CDC), Atlanta, GA or the Center for Vaccine Development, Baltimore, MD have previously been described [18] (link). These strains were Salmonella serovars of various O serogroups (B, C1, C2, D, E1, O28 and O38) and H types (b, c, d, h, i, g, k, l, m, p, s, t, v, y, z10 and z29). Nine O serogroup B, Phase 1 flagella antigen H:i reference strains from the CDC were used to develop a PCR that discriminates between Salmonella Typhimurium and I 4,[5],12:i:- (Table 1). The NTS test strains consist of 327 Salmonella serogroup B and D isolates that were originally obtained from the blood cultures of febrile patients at l'Hôpital Gabriel Touré in Bamako, Mali. These strains were identified by conventional microbiological and classical serotyping methods by the CVD and CDC, as previously described [18] (link); 69 isolates were O serogroup D, including 37 Salmonella Dublin and 32 Salmonella Enteritidis, and 258 isolates were O serogroup B.

We have designed this approach with whole proteome analyses in mind, so as a proof-of-concept, we applied the approach to the study of host adaptation in Salmonella enterica, using S. Enteritidis and S. Gallinarum as a specific example. For a summary of the workflow of this portion of the analysis, see Supplementary Figure 1. Genomes for Salmonella Enteritidis str. P125109 (AM933172.1) and Salmonella Gallinarum str. 287/91 (AM933173.1) were retrieved. Custom gene models were constructed by searching each S. Enteritidis protein coding gene against the Uniref90 database. Profiles were built using sequences showing 40% or greater sequence identity to the original query protein.
In order to assess whether models built from a small number of sequences performed well, we tested the performance of models using proteins from the humsavar database (http://www.uniprot.org/docs/humsavar), which catalogs human polymorphisms and disease variants for a wide range of proteins which have different levels of representation in Uniprot90, and therefore result in profile models built from varying numbers of sequences. We built custom models for each protein in the database, then separated the models into classes based on both number of sequences and effective sequence number. We then computed AUC values for predictions made on variant data from each class. Results from this test can be found in Supplementary Table S2. We saw no classes where custom models performed worse than Pfam models, so we did not filter models built from few sequences. We did, however, use Pfam models for those proteins with no homologs in Uniref90 rather than build a model from the query sequence, as we felt this could bias results towards the reference species.
To assess the quality of our analysis, we wished to compare our results to those of Nuccio and Bäumler (2014) (link), so we used the ortholog calls provided in their supplementary material. Genes represented by a custom model were scored using the appropriate model (n = 3154), all others were scored using Pfam domain models. hmmsearch was used for scoring. If hits to multiple Pfam models occurred, any overlapping hits were competed based on E-value. Orthologs with incompatible Pfam domain architectures (n = 32) were excluded from scoring, but counted as hypothetically attenuated coding sequences (HACs) in Figure 2A if they involved a loss of domain in one serovar. We anticipate that the variance of delta-bitscores will increase with evolutionary distance, so rather than establish a fixed scoring cutoff for HACs, we identify loss-of-function mutations using an empirical distribution. We set a score threshold at which 2.5% of genes on the least dispersed side of the distribution would be called as HACs. If the two bacteria show an equal rate of protein function loss, this would result in 5% of orthologous proteins being called as HACs, however if one proteome shows a greater degree of functional degradation, this will result in a greater proportion of orthologs being classified as HACs.
Functional classification was performed using pathways from the KEGG database (Kanehisa et al., 2016 (link)). We grouped genes into four categories: those present in a pathway but with no ortholog in the other serovar; genes identified as hypothetically disrupted coding sequences (HDCs) by Nuccio and Bäumler (2014) (link); genes identified by our DBS method as HACs, but not as HDCs; and finally genes not identified as non-functional by either method. dN/dS values were calculated using PAML (Yang, 1997 (link)), and for the comparison of DBS and dN/dS, genes were filtered for those with dN > 0 and dS > 0.0001. Correlations between measures were computed using a Spearman’s rho statistic (R package cor).
In our investigation of multiple S. enterica isolates, for all orthologous groups with a gene present in S. Enteritidis, scores were collated, and if individual scores were significantly different to the median score for all isolates, we identified these proteins as HACs. Significant difference was calculated in a similar way to the pairwise comparisons, with the score corresponding to the most extreme 2.5% of delta-bitscores on the least dispersed side of an empirical distribution being used as the cutoff.

Salmonella association and invasion assays were performed according to Zavala et al. (2016) (link). Caco-2/TC-7 cells that were routinely grown in Dulbecco’s modified Eagle’s minimal essential medium (DMEM) (GIBCO BRL Life Technologies Rockville, MD. USA), supplemented with 15% heat-inactivated (30 min at 60°C) fetal bovine serum (FBS, PAA, GE Healthcare Bio-Sciences Corp., USA), 1% non-essential amino acids (GIBCO BRL Life Technologies Rockville, MD. USA), and the following antibiotics (Parafarm, Saporiti SACIFIA, Buenos Aires, Argentina): penicillin (12 UI/mL), streptomycin (12 μg/mL), gentamicin (50 μg/mL). Cells were seeded in 24-well culture plates (Corning, NY, USA) at 2.5 × 10⁵cells per well and incubated at 37°C in a 5% CO₂ — 95% air atmosphere. Caco-2/TC-7 cells were used at post-confluence after 7 days of culture.
Salmonella enteritidis serovar enteritidis CIDCA 101, provided by Dr. H. Lopardo, was grown in nutritive broth (Biokar Diagnostics, Beauvais, France) for 18 h at 37°C (Golowczyc et al., 2007 (link)). Confluent Caco-2/TC-7 monolayers were washed twice with sterile PBS (pH 7.2). Cells were pre incubated for 1 h at 37°C in a 5% CO₂—95% air atmosphere with 250 μL EPS solutions (300, 500 and 800 mg/L in DMEM) or 250 μL DMEM in the case of Salmonella association and invasion controls. Afterwards, 250 μL of Salmonella suspension (1 × 10⁷ CFU/mL) were added to each well and incubated 1 h at 37°C in a 5% CO₂—95% air atmosphere. For association assays, cells were washed three times with PBS and lysed with 500 μL/well of bi-destilled water. The number of associated Salmonella (adhering and invading) was determined by serial dilutions on 0.1% w/v tryptone followed by colony counts on nutrient agar. Salmonella invasion was determined by counting only bacteria located in the Caco-2/TC-7 cells. For this purpose, the monolayer incubated with Salmonella as previously described, were treated with 0.5 mL/well of gentamicin (100 μg/mL PBS) for 1 h at 37°C. Subsequently, cells were lysed and colony counts performed as described above.

Stock cultures of EPS-producing Lacticaseibacillus paracasei CIDCA 8339, CIDCA 83123 and CIDCA 83124 isolated from kefir grains were stored in 12% w/v non-fat milk solids at −80°C. Strains were grown in MRS broth (Difco Laboratories, Detroit, MI, USA) at 20°C (48 h), 30°C (24 h) and 37°C (24 h) under aerobic conditions previous to each experiments (Bengoa et al., 2018a (link)). Salmonella enterica serovar Enteritidis CIDCA 101 (Zavala et al., 2016 (link)) used for association/invasion experiments was grown in nutrient broth (Biokar Diagnostics, Beauvais, France) for 18 h at 37°C.

In this study, 49 Salmonella enterica strains from the Bacterial Strain Collection of the Laboratory of Immunology and Molecular Biology were included, and Salmonella enteritidis (ATCC® 13076™) were used as a positive control. The strains were previously serotyped using the Kauffmann−White scheme and correspond to the serotypes, namely, S. enteritidis (n = 4), S. typhimurium (n = 2), S. braenderup (n = 1), S. newport (n = 1), S. grupensis (n = 1), and S. uganda (n = 1) isolated from cases of gastroenteritis in humans [22 (link)] and S. paratyphi B (n = 24) and S. heidelberg (n = 15) isolated from poultry farms located in the region of Tolima [23 (link)] and Santander [24 (link)].