Ummi Kalthum Azlan, Hadi Azamuddin Abd Hakim, Pei Lou Wong, Noor Hanini Che Lah, Chenyu Zhou, Nur Irlia Sofea Mohammad Zamani, Xiaohui Tong, Rongchun Han, Norazlan Mohmad Misnan, Murni Nazira Sarian, Emelda Rosseleena Rohani, Ahmed Mediani, Faidruz Azura Jam, Hamizah Shahirah Hamezah
Background: leaves might substantially affect the effectiveness of therapy; hence, optimizing the bioactive phytoconstituents is crucial. Liquid chromatography-mass spectrometry(LC-MS)-based metabolomics is essential to (i) evaluate differences in phytochemical composition and bioactivities across leaf maturity, (ii) profile and model leaf metabolome across maturity stages using untargeted ultra-high performance liquid chromatography tandem mass spectrometry (UHPLC-MS/MS) for key metabolites annotation, and with multivariate data analysis (MVDA) to reveal maturity-linked clusters and discriminant metabolites, and (iii) integrate multivariate regression to relate metabolite signatures to bioassays and nominate optimal maturity stages. Methods: leaves were harvested at different maturity stages (day-30, 45, and -60) from Kuala Terengganu, Malaysia. Aqueous ethanolic extracts were subjected to bioactivity assays, including total phenolic (TPC) and flavonoid contents (TFC), 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging, ferric reducing power test (FRAP) antioxidant activity, and acetylcholinesterase (AChE) inhibition. Metabolomics profiling was conducted, and the correlation of phytochemicals to bioactivities was performed by multiple MVDA models. Results: Day-60 leaf extracts exhibited higher TPC and TFC compared to younger stages, with strong antioxidant activity as indicated by elevated DPPH and FRAP values. In contrast, AChE inhibition was highest in day-30 extracts. Metabolomics profiling with chemometrics integration using MVDA identified and tentatively annotated 27 metabolites with predominantly flavonoids (59%), phenolic acids (19%), glucosinolates (7%), and minor components including coumarins, folates, and alkaloid-like compounds. Principal component analysis (PCA) and partial least squares (PLS) biplot revealed distinct metabolic clustering, with day-60 extract formed clear separated cluster, determined by accumulation of quercetin rutinoside, chlorogenic acid, kaempferol derivatives, and procyanidin B2. PLS biplot demonstrated that quercetin derivatives, chlorogenic acid, and procyanidin B2 were positively associated with antioxidant indicator, while coumarin and folic acid aligned with AChE inhibition. Unknown metabolites indicated by chromatogram peak area with high Variable Importance in Projection (VIP) scores (>1) also contributed to bioactivity trends and separation, lead to potential of unannotated phytochemicals. Conclusion: leaf. Chemometrics integration of metabolomics analysis with bioassays shows that maturity stages drive distinct metabolite clustering and bioactivities, with day-60 yielding the highest phenolic, flavonoid, and antioxidant capacity (DPPH/FRAP), while day-30 exhibits the strongest AChE inhibition, thereby, defining maturity-specific optimal harvests.