The prevalence and characterisation of Escherichia coli on fresh produce from selected farms, retail outlets and markets in the Western Cape

[摘要] ENGLISH ABSTRACT: South Africa is a water scarce country and farmers are forced to irrigate crops with river water. Contamination of South African rivers has been reported and the carry-over of bacteria from river water to produce has been confirmed. Foodborne outbreaks linked to fresh produce are increasing world-wide.A total of 151 fresh produce samples (lettuce, tomatoes, beans, peas, coriander, basil, mint, rocket, thyme, spinach, cabbage, parsley and sprouts) were sourced from small-scale and commercial farms, farmers' markets and retail outlets. Total coliforms (TC) and E. coli loads on the produce were determined with Colilert-18. Isolates were phenotypically characterised and identified with the API system and the E. coli identification confirmed with uidA PCR. Sixty-three E. coli isolates were identified. Three were not identified as E. coli with the API system but were positive for the uidA gene.The TC loads for the produce from the farms, farmers' markets and retail outlets were all in the range of log 3 to log 8.38 MPN.100 mL-1. Escherichia coli was found to be most prevalent on produce samples from farmers' markets with the highest E. coli load (log 7.38 MPN.100 mL-1) on cabbage sampled from a commercial farm. Escherichia coli were present on 8% of the produce samples. The maximum TC and E. coli loads found on the fresh produce were log 8.38 and log 7.38 MPN.100 mL-1, respectively. The lowest risk in terms of TC and E. coli presence and load was observed on fresh produce from retail outlets and the highest risk was on fresh produce from farmers' markets.Phenotypic dendrograms and a PCA plot were statistically constructed to determine similarity groupings of the isolates and three main E. coli clusters were formed. These three clusters could not be directly linked to a specific produce type or source type. A larger variation E. coli phenotypes was observed present on fresh produce within the three clusters.All E. coli isolates were also subjected to triplex and multiplex PCR analysis to identify their phylogenetic groups and the presence of INPEC and ExPEC strains. Fourteen isolates belonged to genotypic group A0, 11 to A1, 20 to B1, 7 to B23 and 11 to D2. Thus a large variation E. coli genotypes are present but it cannot be linked to a specific source type or produce type. Multiplex PCR testing for INPEC revealed that none of the E. coli isolates were carriers of the INPEC genes. The isolates were also tested for the presence of ExPEC gene sequences: papA, papC, sfa/foc, iutA, kpsMT II and afa/dra. None of the isolates were classified as ExPEC (which required the presence of two or more genes) but three of the isolates did test positive for the presence of the kpsMT II gene. The latter could indicate that potentially pathogenic E. coli can be evolving in the environment and increase the risk of pathogenic E. coli occurring on fresh produce.In conclusion, the presence of E. coli (commensal or pathogenic) on fresh produce is unacceptable according the South African Department of Health. According to this study the identification of E. coli types could not be correlated with the presence of E. coli on the different produce types and thus the presence of E. coli on fresh produce is unpredictable. It is recommended that extensive safety precautions should be in place throughout every step in the production chain from harvest to the consumer's kitchen to reduce the probability of contamination of fresh produce.

[发布机构] Stellenbosch University