Integrated LC × LC-ASCA workflow for chemically interpretable fingerprinting of cultivar and cultivation dependent...

Cannabis sativa inflorescences exhibit a highly complex metabolome composed of cannabinoids, terpenes, and other metabolites relevant for medicinal and food related applications. However, most analytical studies focus on targeted compound families or single cultivation variables, limiting comprehensive evaluation of genotype and cultivation related chemical variability. These challenges highlight the need for selective analytical methods combined with chemometric tools for metabolomic fingerprinting and identification of chemical markers associated with cultivation variables. An optimized RPLC × RPLC method using Smart Active Modulation was applied to generate high-information-content chromatographic fingerprints comprising 32 terpenes, 10 cannabinoids, and multiple untargeted metabolites from inflorescences of three Cannabis cultivars grown under indoor, outdoor, and greenhouse conditions. ANOVA-Simultaneous Component Analysis (ASCA) was successfully applied to the multivariate datasets obtained from a single 2D-LC analysis, enabling evaluation of cultivar, cultivation method, and their interaction through chemically interpretable fingerprinting (targeted and untargeted compounds). Cultivar emerged as the primary factor influencing chemical composition (24.28% of variability), whereas cultivation method also had a significant, although smaller, effect (4.16% of variability), with outdoor and greenhouse grown inflorescences exhibiting greater similarity to each other than to indoor-grown plants. No significant interaction between factors was observed. In addition, discriminant chemical markers were identified, revealing higher relative abundances of monoterpenes in indoor samples, whereas sesquiterpenes predominated in outdoor and greenhouse samples independently of cultivar. The proposed LC × LC-ASCA workflow demonstrates how comprehensive chromatographic fingerprints can be transformed into chemically interpretable information through factor oriented chemometric decomposition, enabling evaluation of structured metabolic variability in highly complex natural matrices.