Characterization of bacterial communities of donkey milk by high-throughput sequencing.

The interest in donkey milk (DM) is growing because of its functional properties and nutritional value, especially for children with allergies and food intolerances. However, most of the available reports of DM microbiota are based on culture-dependent methods to investigate food safety issues and the presence of lactic acid bacteria (LAB). The aim of this study was to determine the composition of DM bacterial communities using a high-throughput sequencing (HTS) approach. Bulk milk samples from Italian donkey dairy farms from two consecutive years were analysed using the MiSeq Illumina platform. All sample reads were classified into five phyla: Proteobacteria, Firmicutes, Bacteroidetes, Actinobacteria, and Verrucomicrobia. The most prevalent genera-Pseudomonas, Ralstonia, Acinetobacter, Cupriavidus, Citrobacter and Sphingobacterium-were Gram-negative bacteria. The core microbiota was composed of genera that comprise commonly associated milk bacteria, LAB and species normally found in soil, water and plants. Reads assigned to LAB genera-Streptococcus, Lactococcus, Enterococcus, Leuconostoc, Lactobacillus, and Carnobacterium-corresponded on average to 2.55% of the total reads per sample. Among these, the distribution of reads assigned to coccus- and bacillus-shaped LAB was variable between and within the farms, confirming their presence and suggesting a complex population of these bacteria in DM. The present study represents a general snapshot of the DM microbial population, underlining its variability and motivating further studies for the exploitation of the technological potential of bacteria naturally present in DM.