Assessment of pathogenic potential in non-pathogenic industrially relevant bacteria

TL;DR

This study assesses whether industrially relevant non-pathogenic bacteria have pathogenic potential using genomic and phenotypic methods.

Contribution

The study introduces a method to distinguish true virulence factors from niche factors in non-pathogenic bacteria.

Findings

Genomic screening found no true virulence factors in industrially relevant bacterial strains.

Phenotypic assays showed no cytotoxic or hemolytic activity in tested strains.

Clinical and industrial isolates were phylogenetically indistinguishable in core genome analysis.

Abstract

Assessment of the pathogenic potential (virulence and toxicity) in non-pathogenic bacterial species is a challenge as it relies on methods developed for assessment of species known to be pathogenic. Here, we have applied and evaluated some of these methods on industrially relevant bacteria to differentiate between ‘true’ virulence factors applying only to pathogens and niche factors being defined as promoting colonization and survival rather than pathogenicity and as being present also in non-pathogenic bacteria. We examined the pathogenicity of 49 strains from 9 industrially relevant bacterial species (Lactobacillus gasseri, Lactobacillus jensenii, Lactobacillus delbrueckii, Lacticaseibacillus rhamnosus, Limosilactobacillus fermentum, Latilactobacillus curvatus, Ligilactobacillus salivarius, Staphylococcus carnosus and Staphylococcus xylosus), including 14 clinical isolates of the same species, through genomic screening and phenotypically through assays established for pathogenic bacteria. The genomes were screened against the Virulence Factor Database (VFDB), and thresholds (>80% nucleotide or protein identity, >70% coverage) provided by the European Food Safety Authority (EFSA) were adopted to differentiate between genes of potential concern and genes of no concern. Core genome analysis was performed to determine whether the clinical isolates were phylogenetically related to the industrial isolates. The genotypic assessment did not reveal the presence of true virulence factors in the examined strains, and in the core genome analysis, the clinical isolates could not be distinguished from the industrial strains. Furthermore, cytotoxicity toward Vero cells, negative impact on Caco-2 cell viability and haemolytic activity on blood agar plates were examined, and none of the tested strains exhibited any activity in these assays. Overall, the results suggest that VFDB screening with the EFSA thresholds can be used to differentiate between true virulence factors and niche factors. Furthermore, the use of phenotypic assays supports the genotypic assessment, albeit expert knowledge is required to interpret the results.

Figures