Abstract
Out of 2495 samples, L. monocytogenes was isolated from 262 (10.5%). Among these, 30 isolates (11.5% of the 262) exhibited unique phenotypic and genetic characteristics compared to reference strains. Hemolysin-negative L. monocytogenes isolates have been increasingly reported in recent years and are challenging to identify due to their altered phenotypic traits and limitations of standard microbiological methods. This study aimed to evaluate the performance of MALDI-TOF MS in identifying and differentiating 30 hemolysin-negative and hemolysin-positive L. monocytogenes isolates and 12 reference strains, using both a commercial Bruker database and a proprietary in-house database developed from newly characterized isolates. The Bruker database correctly identified only 21% of the environmental isolates, misclassifying most as L. innocua, and showed 83.3% accuracy for reference strains. In contrast, the in-house database achieved 96.6% and 100% accuracy for the environmental and reference strains, respectively. Statistical methods, including hierarchical clustering, heatmaps, PCA, and Pearson correlation, revealed grouping based on phenotypic traits and origin, with key peptides influencing classification. Biomarkers linked to hemolysis and antibiotic resistance differentiated the environmental isolates from reference strains. These findings highlight the need for the development of customized spectral databases to improve the detection of L. monocytogenes in food safety monitoring.