Development of a propidium monoazide–droplet digital polymerase chain reaction assay for strain-specific quantification of Lactiplantibacillus plantarum YZH81 in fermented milk

ABSTRACT: Fermented milk is a widely popular probiotic food. In addition to starter cultures, its production process is often enhanced with specific probiotic strains, such as Bifidobacterium spp., Lactiplantibacillus plantarum, and Lacticaseibacillus rhamnosus, although the particular strains employed vary among different products. Because the efficacy of probiotics is strain-specific, determining the composition and abundance of probiotic strains in fermented milk is critical for the quality assessment and regulation of these products. In this study, L. plantarum YZH81, a probiotic strain previously isolated and preserved by our laboratory, was selected as the research subject. Based on whole-genome sequencing and SNP analysis, the whole-genome sequences of 19 other L. plantarum strains were compared, leading to the design of a pair of strain-specific primers and a minor groove binder-modified probe. Subsequent specificity assays confirmed the capability of these oligonucleotides to effectively distinguish L. plantarum YZH81 from other strains. Accordingly, a propidium monoazide (PMA)-droplet digital PCR (ddPCR) method was developed and optimized to effectively differentiate viable from nonviable cells. Compared with quantitative real-time PCR (qPCR), the ddPCR method offers superior specificity, sensitivity, and interference resistance, with a detection limit as low as 102 cfu/mL, demonstrating higher sensitivity than qPCR. When the target strain YZH81 was quantitatively added to fermented milk, the PMA-ddPCR method showed accurate quantification within the range of 104 to 107 cfu/mL, meeting the requirements for strain-level detection in fermented milk. This strain-specific PMA-ddPCR method provides a rapid, sensitive, and effective solution for quality control and regulatory compliance assessment of probiotic fermented milk, enabling precise quantification of viable target strains even in complex product matrices with high specificity and reproducibility.