INVASION OF HEP-2 CELLS BY STRAINS OF SALMONELLA-TYPHIMURIUM OF DIFFERENT VIRULENCE IN RELATION TO GASTROENTERITIS

Experiments to measure the invasiveness of seven strains of Salmonella typhimurium for HEp-2 cells showed that high inocula (100 bacteria/HEp-2 cell), as used by most workers to synchronise events and to increase the number of bacteria which invade, resulted in recovery of significantly less than 1% of the original inoculum after treatment with gentamicin to kill extracellular bacteria. Also, the cell culture medium became acidic, and microscopic examination of Giemsa-stained monolayers immediately following gentamicin treatment revealed high concentrations of bacteria associated with the cells. Moreover, with bacterium-cell interaction beyond 2 h, many HEp-2 cells became rounded, especially with virulent strains W118 and TML. Thus, the biological significance of the quantitative data was uncertain. The fall in pH and the rounding of HEp-2 cells were prevented by the use of a low (1:1) bacterium:cell ratio; but the recovery of bacteria after treatment with gentamicin was still lower than expected by microscopic examination. After treatment of cells with Triton X-100 to release bacteria, many remained bound to residual cell nuclei. Additional treatment with a rubber policeman, and vigorous pipetting to disperse aggregates of bacteria and cell debris, increased the recovery to c. 10% of the initial inoculum after interaction for 2 h, and 30-80% after 4 h, depending on the strain and experimental conditions. The pattern of invasiveness, but not the absolute count, was highly reproducible on different days and in different hands. However, after interaction exceeding 2 h, the distribution of bacteria was uneven, many cells having no associated organisms, others showing microcolonies. Either this variation does not happen with high inocula, or it is occluded by the high concentration of bacteria associated with the monolayer. Uptake of bacteria depends on the batch of fetal calf serum used in the cell culture medium. The bacterial phenotype is important: bacteria in early or mid log phase entered cells more efficiently, and bacteria grown in Hartley Digest Broth were significantly better at invading HEp-2 cells than those grown in Myosate Broth. Centrifuge-assisted inoculation of HEp-2 cells with bacteria may grossly distort the results, particularly with some avirulent strains.