Global meat consumption is increasing, raising significant environmental and health concerns. The transition to plant-based diets is gaining traction, driven by consumer demand for sustainable and healthier food options. Plant-based meat alternatives (PBMAs) are a key part of this shift, experiencing substantial market growth. However, research on the safety and microbiological properties of PBMAs is limited. This study aims to characterize the microbial communities present in commercially available PBMAs in Austria, investigating potential safety risks and the influence of factors like protein source and texture on microbial composition. The hypothesis is that products within the same protein source and texture group will exhibit more similar microbial communities due to similar processing methods. This research is vital for informing food safety regulations, optimizing shelf life, and addressing sustainability concerns, especially considering that up to 30% of food is lost in primary processing due to microbial spoilage or contamination. Furthermore, limited knowledge about the microbial compositions in our food poses concerns about the positive effects of microbial diversity on human health.

Extensive research exists on the environmental and health benefits of reducing meat consumption and increasing plant-based food intake. Studies have shown the positive impact of plant-based diets on reducing risks of type 2 diabetes, cardiovascular diseases, and metabolic syndrome. Market research indicates a strong increase in the sales and diversity of PBMAs, particularly among flexitarians. However, research on the specific hazards associated with PBMAs, including contaminants like pesticides, mycotoxins, and heavy metals, remains limited. Recent studies have highlighted the presence of significant mycotoxin levels in some PBMAs, emphasizing the need for more comprehensive safety assessments. Microbiological studies on PBMAs, especially those not classified as ready-to-eat, are also scarce. The lack of detailed microbiological data is a major gap hindering the development of appropriate safety regulations and recommendations.

This study analyzed 32 PBMAs purchased from Austrian supermarkets. The samples were categorized into four groups based on protein source (pea or soybean) and texture (minced or fibrous). A combination of culture-dependent and culture-independent methods were used. For culture-dependent analysis, samples were homogenized, and microbial colonies were isolated on various media, followed by 16S rRNA gene Sanger sequencing for identification. Two rounds of cultivation were performed: one using broad-range media and another targeting specific genera identified through sequencing. For culture-independent analysis, DNA was extracted from the samples, and the V3/V4 region of the 16S rRNA gene was amplified and sequenced using Illumina MiSeq. Sequence data were processed using QIIME2 for quality control, denoising, and taxonomic classification. Alpha diversity (Hill-Shannon and Hill-Simpson indices) was calculated to assess the diversity within samples. Beta diversity analysis (Bray-Curtis, Jaccard, and JSD) was performed using t-SNE and PERMANOVA to compare microbial communities between sample groups. LEfSe was used to identify differentially abundant taxa between groups. Furthermore, fifteen potential pathogens and six unclassified Enterobacteriaceae isolates underwent whole-genome sequencing using Oxford Nanopore MinION. Genome analysis involved assembly, annotation, and identification of virulence factors and antimicrobial resistance genes.

The study found that lactic acid bacteria (LAB), specifically *Leuconostoc* and *Latilactobacillus*, were dominant in most (84%) of the PBMAs, with low alpha diversity. *Pseudomonadota* dominated the remaining samples. Culture-dependent methods revealed diverse *Bacillus*, *Enterobacteriaceae*, and potential pathogens. t-SNE analysis revealed three distinct microbial community profiles based on the relative abundances of *Leuconostocaceae*, *Latilactobacillus*, and *Pseudomonadota*. PERMANOVA analysis showed that protein source and texture significantly affected microbial composition, but explained only a small portion of the variance (15.94-23.44%), indicating that the manufacturer played a larger role. LEfSe identified specific genera predictive of product categories. Whole-genome sequencing revealed that several isolates possessed virulence factors and antimicrobial resistance genes (AMRG), including *Escherichia coli* with AIEC-associated virulence factors and AMRG, and various *Klebsiella* species with beta-lactamase genes. *Bacillus paranthracis* isolates harbored non-hemolytic enterotoxin genes. *Staphylococcus aureus* isolates also showed multiple AMRG. LAB isolates showed genes for various metabolites linked to meat spoilage, and *Leuconostoc mesenteroides* exhibited greater capabilities for carbohydrate utilization and amino acid biosynthesis compared to *L. sakei*. The study also found that several genera with high relative abundances in amplicon sequencing could not be isolated through cultivation.

The findings highlight the complexity of the microbial communities in PBMAs and emphasize the need for further research to understand the factors driving their composition. While LAB dominance suggests a potential role in product stability and shelf life, the presence of potential pathogens highlights food safety concerns. The low variance explained by protein source and texture in the PERMANOVA analysis, coupled with the high variance explained by the manufacturer when added to the model (though this analysis was deemed inconclusive due to imbalanced sampling), suggests that production practices play a critical role in shaping the microbial community. The presence of virulence factors and AMRG in some isolates raises concerns about potential foodborne illnesses, emphasizing the importance of stringent food safety practices and adequate cooking instructions for consumers. The relatively high percentage of products with a best-before date, instead of an expiry date also highlights potential consumer hesitancy and misinterpretations regarding spoilage detection and product safety.

This study provides novel insights into the microbial ecology of PBMAs, identifying key microbial communities and potential food safety risks. The dominance of LAB suggests potential roles in spoilage and preservation. The presence of potential pathogens, and AMRG underscores the need for further research focused on raw materials, production processes, and consumer education to enhance food safety and optimize shelf life. Future studies should focus on larger sample sizes, examining the influence of specific production steps, and developing methods for improving the detection of spoilage indicators in these products.

The sample size, while significant, may not fully represent the entire range of PBMAs available. The study is geographically limited to Austrian products. The lack of detailed production process information for each product limited the ability to fully explain the variance in microbial communities. Also, the reliance on solely 16S rRNA gene sequencing might have underestimated the diversity of microbial communities present and not detected other microorganisms.