Molecular characterization and antimicrobial resistance profiles of Salmonella typhimurium isolated between 1995 and 2002 from organs and environments of diseased poultry in South Africa

Abstract

Despite the occurrence of S. Typhimurium infections, little is known on the genetic diversity, virulence characteristics and antimicrobial resistance profiles of poultry S. Typhimurium in South Africa. Therefore, S. Typhimurium (n=141) isolated from organs (n=115) and environments (n=26) of diseased poultry between 1995 and 2002 were screened by PCR for bacteriophages, plasmids and Salmonella pathogenicity islands (SPIs) - encoded virulence genes (virulotyping) which are essential for invasion (invA, sopB, gtgB, sspH1, sopE, spvC, and pefA), survival (sifA, gipA, sodC1, gtgE, mig5, and sspH2) and serum killing (rck, and srgA) of the pathogen in the host. Isolates were also characterized by: pulsed field gel electrophoresis (PFGE) for genetic relatedness, and plasmid profiling (n=43). Furthermore, isolates (n=141) were tested for susceptibility to 16 antimicrobials by disk diffusion and further screened by PCR for the carriage of 27 resistance genes, and integrons. Multi-resistant S. Typhimurium definitive phage type (DT) 104 were determined by disk diffusion and confirmed by PCR. All isolates carried SPIs-encoded genes: invA, sopB, and sifA. Bacteriophages-encoded genes (sspH2, sspH1, sodC1, gtgB, and gtgE) occurred in more than 74.5% of the isolates expect for gipA (57.6%), and sopE (19%). The occurrence of plasmid-encoded genes (pefA, mig5, rck, spvC, and srgA) ranged from 48.2% to 74.5%. Two sample t - test showed that virulence genes: gtgB, spvC, gipA, gtgE, mig5, rck and srgA were more frequent (p ? 0.05) in S. Typhimurium isolates from environments. Virulotyping clustered 141 isolates into 59 virulotypes with 97 isolates clustering in 5 predominant virulotypes while 44 were single isolate virulotypes. PFGE grouped 140 isolates into 55 pulsotypes with 66 isolates clustering in 5 major pulsotypes, 51 isolates clustering in small pulsotypes (containing less than 5 isolates) while 33 were single isolate pulsotypes. Ten plasmid profiles ranging from 2kb to 90kb were observed. The most common plasmid profile contained the 90kb plasmid and was observed in 12/43 isolates. Major virulotypes and plasmid profiles corresponded approximately to pulsotypes and clustered isolates recovered from the same farms or during the same period. Virulotyping and PFGE showed identical discriminatory index (D=0.93). Multidrug resistance (resistance to ? 2 antimicrobials) was observed in 97.2% of isolates. High levels of resistance phenotypes and their respective resistance genes were observed for: streptomycin (94.3%) conferred by ant3'Ia (60.3%) and str (50.4%), sulphonamides (87.2%) conferred by sul1 (66%) and sul3 (31.9%), ciprofloxacin (79.4%) conferred by qnrA (79.4%), tetracycline (61%) conferred by tetB (35.5%) and tetG (28.4%), and cefotaxime (55.3%) conferred by blaSHV (57.4%). Two sample t - test revealed that isolates from poultry organs were more resistant (p?0.05) to ampicillin, amoxicillin clavulanic acid, chloramphenicol, tetracycline and sulfamethoxazole - trimethoprim while isolates collected from poultry environments were more resistant to cephalothin, cefotaxime, ceftazidime, colistin sulphate and nalidixic acid. Using the Kappa statistics, there were agreements ranging from good to perfect between phenotype and genotype. In addition, for every phenotypic resistance recorded, at least one corresponding resistance gene was detected. DT104 strains and class1 integrons were observed in 34.7% and 83% of the isolates respectively. Multi-resistant S. Typhimurium (97.2%) also carried SPIs - encoded virulence genes involved in invasion and survival in the host. In addition, more than 50% of resistant isolates to each of the antimicrobials also carried at least 12 virulence genes: invA, sopB, sifA, sspH2, sspH1, sodC1, gtgB, gtgE, pefA, mig5, spvC, and srgA. A significant number (44.9%) of the DT104 strains that were clustered in the same pulsotype X25 also belonged to virulotype V3a which contained 13 virulence genes: invA, sopB, sifA, sspH2, sspH1, sodC1, gtgB, gtgE, pefA, rck, mig5, spvC, and srgA. Most of isolates that belonged to the same antimicrobial resistance profile (phenotype and genotype) carried at least 8 common virulence genes. In conclusion, these data indicate that S. Typhimurium isolated from diseased poultry carry virulence genes that are usually incriminated in Salmonella human outbreaks. Virulotyping and PFGE showed the same discriminatory index (D=0.93) indicating that virulotyping can be an alternative subtyping method in laboratories where PFGE is not available. Salmonella Typhimurium are also genetically diverse since they were recovered from multiple farms and during a period spanning 8 years. Furthermore, isolates were resistant to multiple antimicrobials used in poultry operations (streptomycin, sulphonamides, and tetracycline) and those used to treat human salmonellosis: ciprofloxacin, and cefotaxime. Multidrug resistant isolates carried most of virulence genes. This relationship between virulence and antimicrobial resistance suggests that the adaptation of isolates against antimicrobial effects may induce expression of virulence factors. The increasing incidence of DT104 threatens the public health since DT104 strains are associated with hospitalizations and deaths in humans. Salmonella Typhimurium carried mobile genetic elements (bacteriophages, integrons and plasmids) which pose a public hazard as they propagate virulence and resistance genes with emerging new pathogenic bacteria as a result. Therefore, monitoring and surveillance of salmonellosis and prudent antimicrobials use need more efforts to ensure animal health and food safety for consumers in South Africa.