Cheese is characterized by a rich and complex microbiota that plays a vital role during both production and ripening, contributing significantly to the safety, quality, and sensory characteristics of the final product. In this context, it is vital to explore the microbiota composition and understand its dynamics and evolution during cheese manufacturing and ripening. Application of high-throughput DNA sequencing technologies have facilitated the more accurate identification of the cheese microbiome, detailed study of its potential functionality, and its contribution to the development of specific organoleptic properties. These technologies include amplicon sequencing, whole-metagenome shotgun sequencing, metatranscriptomics, and, most recently, metabolomics. In recent years, however, the application of multiple meta-omics approaches along with data integration analysis, which was enabled by advanced computational and bioinformatics tools, paved the way to better comprehension of the cheese ripening process, revealing significant associations between the cheese microbiota and metabolites, as well as their impact on cheese flavor and quality.

One of the main lactic acid bacterial species found in the kefir grain ecosystem worldwide is Lactobacillus kefiranofaciens, exhibiting strong auto-aggregation capacity and, therefore, being involved in the mechanism of grain formation. Its occurrence and dominance in kefir grains of various types of milk and geographical origins have been verified by culture-dependent and independent approaches using multiple growth media and regions of the 16S rRNA gene, respectively, highlighting the importance of their combination for its taxonomic identification. L. kefiranofaciens comprises two subspecies, namely kefiranofaciens and kefirgranum, but only the first one is responsible for the production of kefiran, the water-soluble polysaccharide, which is a basic component of the kefir grain and famous for its technological as well as health-promoting properties. L. kefiranofaciens, although very demanding concerning its growth conditions, can be involved in mechanisms affecting intestinal health, immunomodulation, control of blood lipid levels, hypertension, antimicrobial action, and protection against diabetes and tumors. These valuable bio-functional properties place it among the most exquisite candidates for probiotic use as a starter culture in the production of health-beneficial dairy foods, such as the kefir beverage.

Feta is the most renowned protected designation of origin (PDO) white brined cheese produced in Greece. The fine organoleptic characteristics and the quality of Feta rely on, among other factors, its overall microbial ecosystem. In this study, we employed 16S rDNA and internal transcribed spacer (ITS) amplicon sequencing, as well as shotgun metagenomics, to investigate the microbiome of artisanal homemade and industrial Feta cheese samples from different regions of Greece, which has very rarely been investigated. 16S rDNA data suggested the prevalence of the Lactococcus genus in the homemade samples, while Streptococcus and Lactobacillus genera prevailed in the industrial control samples. Species identification deriving from shotgun metagenomics corroborated these findings, as Lactococcus lactis dominated two homemade samples while Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus were found to be dominating one industrial sample. ITS data revealed a complex diversity of the yeast population among the samples analyzed. Debaryomyces, Kluyveromyces, Cutaneotrichosporon, Pichia, Candida, and Rhodotorula were the major genera identified, which were distributed in a rather arbitrary manner among the different samples. Furthermore, a number of potential metagenome-assembled genomes (MAGs) could be detected among assembled shotgun bins. The overall analysis of the shotgun metagenomics supported the presence of different foodborne pathogens in homemade samples (e.g., Staphylococcus aureus, Listeria monocytogenes, Enterobacter cloacae, and Streptococcus suis), but with low to very low abundances. Concluding, the combination of both amplicon sequencing and shotgun metagenomics allowed us to obtain an in-depth profile of the artisanal homemade Feta cheese microbiome.

In the present study we investigated the incidence of bacteriocins produced by 236 lactic acid bacteria (LAB) food isolates against pathogenic or opportunistic pathogenic oral bacteria. This set of LAB contained several strains (≥17%) producing bacteriocins active against food-related bacteria. Interestingly only Streptococcus macedonicus ACA-DC 198 was able to inhibit the growth of Streptococcus oralis, Streptococcus sanguinis and Streptococcus gordonii, while Lactobacillus fermentum ACA-DC 179 and Lactobacillus plantarun ACA-DC 269 produced bacteriocins solely against Streptococcus oralis. Thus, the percentage of strains that were found to produce bacteriocins against oral bacteria was ~1.3%. The rarity of bacteriocins active against oral LAB pathogens produced by food-related LAB was unexpected given their close phylogenetic relationship. Nevertheless, when tested in inhibition assays, the potency of the bacteriocin(s) of S. macedonicus ACA-DC 198 against the three oral streptococci was high. Fourier-transform infrared spectroscopy combined with principal component analysis revealed that exposure of the target cells to the antimicrobial compounds caused major alterations of key cellular constituents. Our findings indicate that bacteriocins produced by food-related LAB against oral LAB may be rare, but deserve further investigation since, when discovered, they can be effective antimicrobials.

Staka is a traditional Greek sour cream made mostly from spontaneously fermented sheep milk or a mixture of sheep and goat milk. At the industrial scale, cream separators and starter cultures may also be used. Staka is sometimes cooked with flour to absorb most of the fat. In this study, we employed culture-based techniques, amplicon sequencing, and shotgun metagenomics to analyze the Staka microbiome for the first time. The samples were dominated by Lactococcus or Leuconostoc spp. Most other bacteria were lactic acid bacteria (LAB) from the Streptococcus and Enterococcus genera or Gram-negative bacteria from the Buttiauxella, Pseudomonas, Enterobacter, Escherichia-Shigella, and Hafnia genera. Debaryomyces, Kluyveromyces, or Alternaria were the most prevalent genera in the samples, followed by other yeasts and molds like Saccharomyces, Penicillium, Aspergillus, Stemphylium, Coniospotium, or Cladosporium spp. Shotgun metagenomics allowed the species-level identification of Lactococcus lactis, Lactococcus raffinolactis, Streptococcus thermophilus, Streptococcus gallolyticus, Escherichia coli, Hafnia alvei, Streptococcus parauberis, and Enterococcus durans. Binning of assembled shotgun reads followed by recruitment plot analysis of single reads could determine near-complete metagenome assembled genomes (MAGs). Culture-dependent and culture-independent analyses were in overall agreement with some distinct differences. For example, lactococci could not be isolated, presumably because they had entered a viable but not culturable (VBNC) state or because they were dead. Finally, several LAB, Hafnia paralvei, and Pseudomonas spp. isolates exhibited antimicrobial activities against oral or other pathogenic streptococci, and certain spoilage and pathogenic bacteria establishing their potential role in food bio-protection or new biomedical applications. Our study may pave the way for additional studies concerning artisanal sour creams to better understand the factors affecting their production and the quality.

The microbial terroir of Greek olive varieties remains underexplored. In this study, 62 samples of olive fruits, collected across the harvest period 2019-2020, were analyzed by high-throughput sequencing. The samples represented 38 olive varieties collected from geographically well distributed regions of Greece. Analysis of the bacterial composition revealed that the geographical area was a significant factor in discriminating samples. The core microbiota included Erwinia, Pseudomonas, and members of the Enterobacteriaceae family. Furthermore, a notable variation in bacterial taxa abundances associated with the geographic location was observed. The sampling area was a key discriminant factor for the mycobiota, and the core mycobiota comprised Alternaria, Taphrina, Candida, Wickerhamomyces anomalus and Penicillium. Finally, Redundancy Analysis (RDA) revealed a notable association between environmental characteristics and microbial composition. Specifically, tree age was associated with certain bacterial and fungal taxa (Pearson's correlation p-value adj.[FDR] < 0.05).