Detection and characterization of Campylobacter in air samples from poultry houses using shot-gun metagenomics - a pilot study

Abstract

Background: Foodborne pathogens such as Campylobacter jejuni are responsible for a large proportion of the gastrointestinal infections worldwide associated with poultry meat. Campylobacter spp. can be found in the chicken fecal microbiome and can contaminate poultry meat during the slaughter process. Commonly used sampling methods to detect Campylobacter spp. at poultry farms use fecal droppings or boot swabs in combination with conventional culture techniques or PCR. In this pilot study, we have used air filtering and filters spiked with mock communities in combination with shotgun metagenomics to detect Campylobacter and test the applicability of this approach for the detection and characterization of foodborne pathogens. To the best of our knowledge is this the first study that combines air filtering with shotgun metagenomic sequencing for detection and characterization of Campylobacter.

Results: Analysis of air filters spiked with different levels of Campylobacter, into a background of mock or poultry house communities, indicated that we could detect as little as 200 colony forming units (CFU) Campylobacter per sample using our protocols. The results indicate that even with limited sequencing effort we could detect Campylobacter in the samples analysed in this study. We observed significant amounts of Campylobacter in real-life samples from poultry houses using both real-time PCR as well as shotgun metagenomics, suggesting that the flocks in both houses were infected with Campylobacter spp. Interestingly, in both houses we find diverse microbial communities present in the indoor air which reflect the fecal microbiome of poultry. Some of the identified genera such as Staphylococcus, Escherichia and Pseudomonas are known to contain opportunistic pathogenic species.

Conclusions: These results show that air sampling of poultry houses in combination with shotgun metagenomics can detect and identify Campylobacter spp. present at low levels. This is important since early detection of Campylobacter enables measures to be put in place to ensure the safety of broiler products, animal health and public health. This approach has the potential to detect any pathogen present in poultry house air.

Keywords: Campylobacter spp.; Gelatin air filter; Metagenomics; Microbial communities; Mock communities; Spike controls.

Fig. 1

Design of the experiment using different air filters. The MOCK samples were spiked with 200 (S1) or 20,000 (S2) CFU / half filter of C. jejuni CCUG 11284T. The HOUSE samples were spiked with three different levels (CFU / quarter filter) of C. jejuni 927

Fig. 2

Relative abundances of bacterial genera found in MOCK and HOUSE samples. The first column shows the expected abundances of MOCK community taxa as provided by the manufacturer. Genera with a relative abundance equal or higher than 0.5% in at least one sample were kept for visualization. Taxa relative abundances below 0.01% are not shown

Fig. 3

Campylobacter spp. abundances in MOCK and HOUSE samples using two different methods. A Normalized abundances of Campylobacter spp. as found with Kraken 2 read classification. Read abundances for the different taxa are shown as classified reads per million reads. B Normalized abundances of reads mapped by BBsplit to the genome sequences of C. jejuni CCUG 11284T (MOCK sample spike) or C. jejuni 927 (HOUSE sample spike). Note that in the HOUSE samples we find reads mapping to C. jejuni CCUG 11284T, indicating the presence of strains closely related to this isolate