Sampling and identification of the dominant microbial population of presumptive biofilms in the food processing industry

Introduction
A lot of bacteria are able to attach and grow on almost every surface forming complex microbial communities termed biofilms. Since they are highly resistant to multiple stress conditions and disinfectants, biofilms have become a concerned source of contamination in the industry including food processing plants. The importance and role of biofilms in persistent infections with for example spoilage organisms and pathogenic bacteria is however insufficiently known. Sampling, detection and characterization of biofilms in the food industry is fundamental. The aim of this study was to evaluate different surface sampling methods to assess the presence of biofilms in different food processing companies and to identify the microbial population of these presumptive biofilms.
Material and methods
In 8 different food processing companies (producing oven foods, dairy products, meat products, baker’s yeast, sauces and egg products) surfaces were sampled after cleaning and disinfection using 2 methods. The first method was scraping with a cell scraper(VWR) followed by swabbing with a floqswab(Copan). Simultaneous a second method, swabbing using a spongestick(3M) hydrated with 10ml of ¼ ringers solution(Oxoid), was performed on a nearby surface. Sampled areas differed from 20cm² to 1200cm² but were similar between the two sampling methods. Different microbiological analyses were performed on both types of samples. After enumeration, morphologically different colonies were isolated, purified and stored in glycerol tubes at -80°C to collect the dominant flora. The dominant bacteria were identified using (GTG)5 clustering followed by 16S rRNA gene sequencing. Isolates were identified using the EzTaxon database.
Results
For total aerobic plate count (TAC, n=133) slightly more points were found to be contaminated using the spongestick (59.4%) in comparison with the scraper/floqswab (54.9%). Beside, average count for TAC on contaminated areas was 2.53±1.68 and 2.01±1.50 log CFU/100cm² for spongestick and scraper/floqswab respectively. For Pseudomonas spp. (n=86), 22.9% of the points were found contaminated using the spongestick with an average count of 2.96±1.85 log CFU/100cm² while 17.4% with the scraper/floqswab providing an average count of 2.50±1.69 log CFU/100cm². The same trend of more positive points and higher counts using spongesticks was observed for Enterobacteriaceae (n=89) and mesophilic spores (n=49). For lactic acid bacteria (LAB, n=80) also more positive points were observed with the spongestick but the average count of both sampling methods was comparable.
In a meat processing company the microbial population in presumptive biofilms was dominated by Gram-negative bacteria mainly Pseudomonas spp. Also Psychrobacter spp., Klebsiella spp., Acidovorax spp. and Stenotrophomonas spp. were identified. In a sauce producing company again Gram-negative bacteria dominated: mostly Pseudomonas spp. but also Acinetobacter spp., Shewanella spp., Enterobacter spp., Lelliottia spp., Stenotrophomonas spp and Citrobacter spp.. Characterization of the isolates from the other 6 food companies showed a wide diversity in the dominant flora with both gram-positive and gram-negative micro-organisms.
Discussion and conclusion
When comparing both sampling methods for all investigated microbiological parameters, sampling with the spongestick provided not only more contaminated points but also higher counts. Only for LAB the average count was similar for both sampling methods. Characterization of the presumptive biofilms showed a wide diversity of micro-organisms within the same company and even sampling point. Between the different companies both large differences as similarities (e.g. Pseudomonas spp.) were found. Further characterization based on the biofilm-forming ability of the isolates is planned.