Application of dual-marker DNA metabarcoding to verify label accuracy in plant-based protein foods.

The global market for plant-based protein foods is rapidly expanding; however, complex supply chains and multi-step processing pose substantial traceability challenges. This study applied DNA metabarcoding to identify plant-derived ingredients in plant-based protein foods, involving the amplification and sequencing of short DNA marker fragments. To enhance the efficiency of species detection, a dual-marker approach using ITS2 and trnL(UAA) was applied to analyze a mock mixed library containing 37 confirmed species. Following database comparison and selection, the NCBI nucleotide database was adopted for ITS2 and a custom database of 37 plant species for trnL(UAA). To verify the accuracy of product labeling, 48 commercial plant-based protein foods in China were tested, including 26 plant-based meat alternatives (PBMAs) and 22 plant-based dairy alternatives (PBDAs). All labeled plant ingredients were successfully detected across the samples. However, 35.42% of products contained one or more undeclared species. The non-compliance rate was significantly higher in PBMAs (57.7%, 15/26) than in PBDAs (9.1%, 2/22). This discrepancy aligns with the more complex formulations, multi-step processing, and longer supply chains characteristic of the PBMAs category. The study confirms the robustness of dual-marker DNA metabarcoding in detecting both declared and undeclared plant components in processed foods. These results highlight the need for improved quality control and regulatory strategies to ensure labeling compliance in plant-based protein foods.