Abstract
Pak choi exhibits a wide range of phenotypic and morphological variations, significantly impacting its carbohydrate composition. This study aimed to analyze these variations by employing UPLC-MS/MS technology on eight biological replicates of seven Pak choi cultivars. The untargeted metabolic analysis identified 513 metabolites, focusing on 16 key carbohydrates, including monosaccharides, disaccharides, and polysaccharides. Monosaccharides were the most prevalent, which were followed by di-, poly-, and oligosaccharides. Suzhouqing had the highest number of differentially accumulated metabolites (DAMs), while Xiangqingcai had the least. Notably, the cultivars Xiangqingcai, Suzhouqing, and Aijiaohuang showed significant metabolite differentiation. The study found 114 metabolites that differed significantly between Suzhouqing and Aijiaohuang, of which 69 were upregulated and 45 were downregulated. In Xiangqingcai and Aijiaohuang, 66 metabolites were upregulated and 49 were downregulated. Between Xiangqingcai and Suzhouqing, 80 metabolites were downregulated and 53 were upregulated. Key carbohydrate digestion and absorption pathways were identified alongside the most enriched flavonoid biosynthesis pathway in Xiangqingcai and Suzhouqing. The findings highlight the considerable carbohydrate variation among Pak choi cultivars, providing valuable insights for targeted carbohydrate extraction and improving nutritional and agricultural practices.