A new approach to the study of recovery times of single heat-injured Salmonella cells is described. It comprises the generation of a standard heat-injured culture, serial dilution of this culture to near extinction, inoculation of the serial dilutions across many microtitre plates and measurement of the subsequent recovery and growth using an automated turbidometric analyser. Lag times for individual cells were estimated from turbidity data using a model that accurately extrapolated the growth curve back to the starting inoculum level. Lag times were compared using a number of different commercially available pre-enrichment media. The most typical result was a very broad distribution of lag times at the single cell inoculum level, with many values in excess of 20 h. Even at an inoculum level 10-fold higher, lag times for some injured cells were estimated to be ×10 h. More significantly, it was found that some media recovered more injured cells than others and vice versa. Between the worst and best media there were as many as 3 log 10 cycles difference in the number of cells recoverable. No trends were apparent linking choice of medium with performance. The implications of these findings, in relation to traditional and rapid methodology, are discussed.

The use and performance of an improved diagnostic and selective medium, PEMBA (polymyxin pyruvate egg yolk mannitol bromothymol blue agar), for the detection of Bacillus cereus in foods is described. The distinct colonial appearance of B. cereus on PEMBA permitted the recognition of both strains: those that do precipitate egg yolk and those that do not react with egg yolk. A staining procedure, used to demonstrate microscopically both the presence of lipid globules in vegetative cells and spore morphology of isolates, proved a rapid and reliable confirmatory test which gave complete agreement with a battery of biochemical tests used for this purpose. The quantitative recovery of B. cereus on PEMBA from 143 food samples was not significantly different from counts on KG (Kim and Goepfert), MYP (mannitol egg yolk phenol red), and McClung's media, and the selectivity of PEMBA was generally superior.

The use of sorbitol MacConkey agar (SMAC) performed poorly in supporting growth of stressed Escherichia coli O157:H7 cells. Up to a 3-log difference was observed between counts on SMAC and tryptone soy agar (TSA). It is critical in the risk assessment of certain foods to be able to enumerate stressed and healthy E. coli O157:H7 in a background of potentially healthy competing bacteria. Investigations carried out to overcome the inhibitory effect of SMAC included the reduction of the selective agent concentration, inclusion of a recovery stage in broth prior to plating out, addition of recovery agents, and delayed exposure to the selective agent. The only successful approach was delayed exposure to the selective agent. This was achieved by resuscitating the stressed cells on a membrane placed on the surface of a TSA plate and, after a defined time period sufficient for full resuscitation, transferring the membrane to the surface of a SMAC plate. The choice of membrane material was critical for maintaining the positive sorbitol color change used to identify wild-type E. coli. Track-etched polycarbonate membranes allowed the typical color reactions to be visualized, whereas cellulose acetate did not. The method was validated with E. coli O157:H7 cells stressed by low pH and high salt conditions, whereby all cells that would previously be undetectable on direct inoculation of SMAC were countable.

A new method for detecting salmonellas in foods within 42 h is described. This highly specific and sensitive selective motility procedure gave an efficiency of 96.8% when tested against traditional methods using more than 800 food samples. It is stable at ambient temperatures and can be prepared for use in 5 min.

Three ELISA's for the detection of staphylococcal enterotoxin A (SEA) were evaluated by a collaborative test in five laboratories for possible use in quality control laboratories. Two ELISA's gave quantitative results using polyclonal antibodies (PCA) or monoclonal antibodies (MCA); the third was the commercially available qualitative FEY ELISA test using PCA. Test samples comprised ripened and unripened sausage alone or spiked with 10 μg or 1 μg SEA/100 g, the latter representing a minimum emetic dose. The quantitative MCA ELISA gave more reliable and sensitive results with lower non-specific background responses, lower standard deviations and coefficients of variation than the PCA ELISA; however, the recovery of enterotoxin concentrations from extracts was lower, which was more distinct in unripened than in ripened sausage. The MCA ELISA gave lower values of absolute response but more reliable results for extracts from sausage spiked with 1 μg SEA/100 g. The FEY ELISA gave sufficient reliable results in our experiments with sausage extracts and looks adequate for controlling suspicious raw materials and food products.