Emma Allen-Vercoe

Fimbriae- and flagella-mediated association with and invasion of cultured epithelial cells by Salmonella enteritidis

Salmonella enteritidis expresses flagella and several finely regulated fimbriae, including SEF14, SEF17 and SEF21 (type 1). A panel of mutants was prepared in three strains of S. enteritidis to elucidate the role of these surface appendages in the association with and invasion of cultured epithelial cells. In all assays, the naturally occurring regulatory-defective strain 27655R associated with tissue culture cells significantly more than wild-type progenitor strains LA5 and S1400/94. Compared with wild-type strains, SEF14 mutants had no effect on association and invasion, whereas SEF17, SEF21 and aflagellate mutants showed significant reductions in both processes. Histological examination suggested a role for SEF17 in localized, aggregative adherence, which could be specifically blocked by anti-SEF17 sera and purified SEF17 fimbriae. SEF21-mediated association was neutralized by mannose and a specific monoclonal antibody, although to observe enhanced association it was necessary for the bacteria to be in fimbriate phase prior to infection. Additionally, aflagellate mutants associated and invaded less than motile bacteria. This study demonstrated the potential for multifactorial association and invasion of epithelial cells which involved SEF17 and SEF21 fimbriae, and flagella-mediated motility.

The role of flagella, but not fimbriae, in the adherence of Salmonella enterica serotype Enteritidis to chick gut explant

To gain an understanding of the role of fimbriae and flagella in the adherence and colonisation of Salmonella enterica serotype Enteritidis in chickens, an in-vitro gut adherence assay was developed and used to assess the adherence of a wild-type Enteritidis strain and isogenic non-fimbriate and non-flagellate mutant strains. Enteritidis strain S1400/94, a clinical isolate virulent in chickens, was shown to possess genes which encoded type 1, SEF14, SEF17, plasmid-encoded and long polar fimbriae. Mutant strains unable to elaborate these fimbriae were created by allelic exchange. Each fimbrial operon was inactivated by the insertion of an antibiotic resistance gene cassette. In addition, fliC, motAB and cheA loci, which encode the major subunit of the flagellum, the energy-translation system for motility and one of the chemotaxis signalling proteins, respectively, were similarly inactivated. Non-flagellate mutant strains were significantly less adherent than the wild-type strain, whereas mutant strains defective for the elaboration of any of the types of fimbriae adhered as well as the wild-type strain. A flagellate but non-motile (paralysed) mutant strain and a smooth-swimming chemotaxis-deficient mutant strain were shown to be less adherent than the wild-type strain, but that observation depended on the assay conditions used.

Colonisation of the chicken caecum by afimbriate and aflagellate derivatives of Salmonella enterica serotype Enteritidis

A semi-quantitative cloacal-swab method was used as an indirect measure of caecal colonisation of one-day old and five-day old chicks after oral dosing with wild-type Salmonella enterica serovar Enteritidis PT4 and genetically defined isogenic derivatives lacking the ability to elaborate flagella or fimbriae. Birds of both ages were readily and persistently colonised by all strains although there was a decline in shedding by the older birds after about 21 days. There were no significant differences in shedding of wild-type or mutants in single-dose experiments. In competition experiments, in which five-day old birds were dosed orally with wild-type and mutants together, shedding of non-motile derivatives was significantly lower than wild-type. At 35 days post infection, birds were sacrificed and direct counts of mutants and wild-type from each caecum were determined. Whilst there appeared to be poor correlation between direct counts and the indirect swab method, the overall trends shown by these methods of assessment indicated that flagella and not fimbriae were important in caecal colonisation in these models.

Adhesion of Salmonella enterica var Enteritidis strains lacking fimbriae and flagella to rat ileal explants cultured at the air interface or submerged in tissue culture medium

Rat ileal air interface and submerged explant models were developed and used to compare the adhesion of Salmonella enterica var Enteritidis wild-type strains with that of their isogenic single and multiple deletion mutants. The modified strains studied were defective for fimbriae, flagella, motility or chemotaxis and binding was assessed on tissues with and without an intact mucus layer. A multiple afimbriate/aflagellate (fim-/fla-) strain, a fimbriate but aflagellate (fla-) strain and a fimbriate/flagellate but non-motile (mot-) strain bound significantly less extensively to the explants than the corresponding wild-type strains. With the submerged explant model this difference was evident in tissues with or without a mucus layer, whereas in the air interface model it was observed only in tissues with an intact mucus layer. A smooth swimming chemotaxis-defective (che-) strain and single or multiple afimbriate strains bound to explants as well as their corresponding wild-type strain. This suggests that under the present experimental conditions fimbriae were not essential for attachment of S. enterica var Enteritidis to rat ileal explants. However, the possession of active flagella did appear to be an important factor in enabling salmonellae to penetrate the gastrointestinal mucus layer and attach specifically to epithelial cells.

Distribution, gene sequence and expression in vivo of the plasmid encoded fimbrial antigen of Salmonella serotype Enteritidis.

The pefA gene which encoded the serotype associated plasmid (SAP) mediated fimbrial major subunit antigen of Salmonella enterica serotype Typhimurium shared genetic identity with 128 of 706 salmonella isolates as demonstrated by dot (colony) hybridization. Seventy-seven of 113 isolates of Typhimurium and individual isolates of serotypes Bovis-morbificans, Cholerae-suis and Enteritidis phage type 9b hybridized pefA strongly, whereas 48 isolates of Enteritidis hybridized pefA weakly and one Enteritidis isolate of phage type 14b failed to hybridize. Individual isolates of 294 serotypes and 247 individual isolates of serotype Dublin did not hybridize pefA. Southern hybridization of plasmids extracted from Enteritidis demonstrated that the pefA gene probe hybridized strongly an atypical SAP of 80 kb in size harboured by one Enteritidis isolate of phage-type 9b, whereas the typical SAP of 58 kb in size harboured by 48 Enteritidis isolates hybridized weakly. One Enteritidis isolate of phage type 14b which failed to hybridize pefA in dot (colony) hybridization experiments was demonstrated to be plasmid free. A cosmid library of Enteritidis phage type 4 expressed in Escherichia coli K12 was screened by hybridization for the presence of pef sequences. Recombinant clones which were deduced to harbour the entire pef operon elaborated a PEF-like fimbrial structure at the cell surface. The PEF-like fimbrial antigen was purified from one cosmid clone and used in western blot experiments with sera from chickens infected with Enteritidis phage-type 4. Seroconversion to the fimbrial antigen was observed which indicated that the Enteritidis PEF-like fimbrial structure was expressed at some stage during infection. Nucleotide sequence analysis demonstrated that the pefA alleles of Typhimurium and Enteritidis phage-type 4 shared 76% DNA nucleotide and 82% deduced amino acid sequence identity.

Expression of type 1 fimbriae (SEF 21) of Salmonella enterica serotype enteritidis in the early colonisation of the rat intestine.

The involvement of type 1 fimbriae in colonisation of the rat gastrointestinal tract in vivo was investigated with Salmonella enterica serotype Enteritidis LA5 and a mutant of LA5 denoted EAV3 unable to elaborate type 1 fimbriae (SEF 21). Rats were given a single dose of LA5 or EAV3 or a 1:1 mixture of both. LA5 was found in higher numbers in the stomach and small intestine than EAV3 at 6 h after infection with a single strain, but not after 6 days. LA5 did not out-compete EAV3 when the strains were administered together. Indeed, after 6 and 21 days, EAV3 was found in the distal small intestine and large intestine in far higher numbers than LA5. These findings suggest that SEF 21 have an important role(s) in the early stages of infection in vivo. However, SEF 21 expression may disadvantage the pathogen in the longer term as indicated by EAV3 out-competing LA5 in the gut at 21 days.

Expression of SEF17 fimbriae by Salmonella enteritidis.

M. DIBB‐FULLER, E. ALLEN‐VERCOE, M. J. WOODWARD AND C. J. THORNS. 1997. Specific immunological reagents were used to investigate the expression of SEF17 fimbriae by cultured strains of Salmonella enteritidis. Most strains of Salm. enteritidis tested expressed SEF17 when cultured at temperatures of 18–30°C. However, two wild‐type strains produced SEF17 when also grown at 37 °C and 42 °C. Colonization factor antigen agar was the optimum medium for SEF17 expression, whereas Drigalski and Sensitest agars poorly supported SEF17 production. Very fine fimbriae produced by a strain of Salm. typhimurium were specifically and strongly labelled by SEF17 monoclonal and polyclonal antibodies, indicating considerable antigenic conservation between the two. Curli fimbriae from Escherichia coli were similarly labelled. The production of these fimbriae corellated with the binding of fibronectin by the organism. Congo red binding by cultured bacteria was not a reliable criterion for the expression of SEF17 fimbriae.